search games and optimal kakeya sets yuval peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh yuval peres...
TRANSCRIPT
![Page 1: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/1.jpg)
Search Games and Optimal Kakeya Sets
Yuval Peres 1
Based on joint work with
Y. Babichenko, R. Peretz, P. Sousi and P. Winkler
1Microsoft ResearchYuval Peres Search Games and Optimal Kakeya Sets
![Page 2: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/2.jpg)
Kakeya sets – History
A subset S ⊆ R2 is called a Kakeya set if it contains a unit segmentin every direction.
Kakeya’s question (1917): Is the three-pointed deltoid shape aKakeya set of minimal area?
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 3: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/3.jpg)
Kakeya sets – History
A subset S ⊆ R2 is called a Kakeya set if it contains a unit segmentin every direction.
Kakeya’s question (1917): Is the three-pointed deltoid shape aKakeya set of minimal area?
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 4: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/4.jpg)
Besicovitch and Schoenberg’s constructions
Besicovitch (1919) gave the first deterministic construction of a Kakeya setof zero area.
He also constructed sets of arbitrarily small area where we can rotate a unitsegment.
Besicovitch’s construction was later simplified by Perron and Schoenbergwho gave a construction of a Kakeya set consisting of 4n triangles of area oforder 1/ log n.
n = 1 n = 2 n = 4 n = 256
(Figures due to Terry Tao)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 5: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/5.jpg)
Besicovitch and Schoenberg’s constructions
Besicovitch (1919) gave the first deterministic construction of a Kakeya setof zero area.
He also constructed sets of arbitrarily small area where we can rotate a unitsegment.
Besicovitch’s construction was later simplified by Perron and Schoenbergwho gave a construction of a Kakeya set consisting of 4n triangles of area oforder 1/ log n.
n = 1 n = 2 n = 4 n = 256
(Figures due to Terry Tao)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 6: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/6.jpg)
Besicovitch and Schoenberg’s constructions
Besicovitch (1919) gave the first deterministic construction of a Kakeya setof zero area.
He also constructed sets of arbitrarily small area where we can rotate a unitsegment.
Besicovitch’s construction was later simplified by Perron and Schoenbergwho gave a construction of a Kakeya set consisting of 4n triangles of area oforder 1/ log n.
n = 1 n = 2 n = 4 n = 256
(Figures due to Terry Tao)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 7: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/7.jpg)
Besicovitch and Schoenberg’s constructions
Besicovitch (1919) gave the first deterministic construction of a Kakeya setof zero area.
He also constructed sets of arbitrarily small area where we can rotate a unitsegment.
Besicovitch’s construction was later simplified by Perron and Schoenbergwho gave a construction of a Kakeya set consisting of 4n triangles of area oforder 1/ log n.
n = 1
n = 2 n = 4 n = 256
(Figures due to Terry Tao)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 8: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/8.jpg)
Besicovitch and Schoenberg’s constructions
Besicovitch (1919) gave the first deterministic construction of a Kakeya setof zero area.
He also constructed sets of arbitrarily small area where we can rotate a unitsegment.
Besicovitch’s construction was later simplified by Perron and Schoenbergwho gave a construction of a Kakeya set consisting of 4n triangles of area oforder 1/ log n.
n = 1 n = 2
n = 4 n = 256
(Figures due to Terry Tao)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 9: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/9.jpg)
Besicovitch and Schoenberg’s constructions
Besicovitch (1919) gave the first deterministic construction of a Kakeya setof zero area.
He also constructed sets of arbitrarily small area where we can rotate a unitsegment.
Besicovitch’s construction was later simplified by Perron and Schoenbergwho gave a construction of a Kakeya set consisting of 4n triangles of area oforder 1/ log n.
n = 1 n = 2 n = 4
n = 256
(Figures due to Terry Tao)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 10: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/10.jpg)
Besicovitch and Schoenberg’s constructions
Besicovitch (1919) gave the first deterministic construction of a Kakeya setof zero area.
He also constructed sets of arbitrarily small area where we can rotate a unitsegment.
Besicovitch’s construction was later simplified by Perron and Schoenbergwho gave a construction of a Kakeya set consisting of 4n triangles of area oforder 1/ log n.
n = 1 n = 2 n = 4 n = 256
(Figures due to Terry Tao)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 11: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/11.jpg)
Besicovitch and Schoenberg’s constructions
Besicovitch (1919) gave the first deterministic construction of a Kakeya setof zero area.
He also constructed sets of arbitrarily small area where we can rotate a unitsegment.
Besicovitch’s construction was later simplified by Perron and Schoenbergwho gave a construction of a Kakeya set consisting of 4n triangles of area oforder 1/ log n.
n = 1 n = 2 n = 4 n = 256
(Figures due to Terry Tao)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 12: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/12.jpg)
New connection to game theory and probability
In this talk we will see a probabilistic construction of an optimal Kakeya setconsisting of triangles.
We do so by relating these sets to a game of pursuit on the cycle Zn
introduced by Adler et al.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 13: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/13.jpg)
New connection to game theory and probability
In this talk we will see a probabilistic construction of an optimal Kakeya setconsisting of triangles.
We do so by relating these sets to a game of pursuit on the cycle Zn
introduced by Adler et al.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 14: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/14.jpg)
New connection to game theory and probability
In this talk we will see a probabilistic construction of an optimal Kakeya setconsisting of triangles.
We do so by relating these sets to a game of pursuit on the cycle Zn
introduced by Adler et al.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 15: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/15.jpg)
A. S. Besicovitch.On Kakeya’s problem and a similar one.Math. Z., 27(1):312–320, 1928.
Roy O. Davies.Some remarks on the Kakeya problem.Proc. Cambridge Philos. Soc., 69:417–421, 1971.
Micah Adler, Harald Racke, Naveen Sivadasan, Christian Sohler,and Berthold Vocking.Randomized pursuit-evasion in graphs.Combin. Probab. Comput., 12(3):225–244, 2003.
Yakov Babichenko, Yuval Peres, Ron Peretz, Perla Sousi, andPeter Winkler.Hunter, Cauchy Rabbit and Optimal Kakeya Sets.Transactions AMS, to appear; arXiv:1207.6389
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 16: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/16.jpg)
Definition of the game Gn
Two players
Hunter Rabbit
Where?
On Zn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 17: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/17.jpg)
Definition of the game Gn
Two players
Hunter Rabbit
Where?
On Zn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 18: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/18.jpg)
Definition of the game Gn
Two players
Hunter
Rabbit
Where?
On Zn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 19: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/19.jpg)
Definition of the game Gn
Two players
Hunter
Rabbit
Where?
On Zn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 20: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/20.jpg)
Definition of the game Gn
Two players
Hunter Rabbit
Where?
On Zn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 21: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/21.jpg)
Definition of the game Gn
Two players
Hunter Rabbit
Where?
On Zn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 22: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/22.jpg)
Definition of the game Gn
Two players
Hunter Rabbit
Where?
On Zn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 23: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/23.jpg)
Definition of the game
When?
At night – they cannot see each other....
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 24: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/24.jpg)
Definition of the game
When?
At night – they cannot see each other....
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 25: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/25.jpg)
Definition of the game
When?
At night – they cannot see each other....Yuval Peres Search Games and Optimal Kakeya Sets
![Page 26: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/26.jpg)
Definition of the game Gn
Rules
At time 0 both hunter and rabbit choose initial positions.
At each subsequent step, the hunter either moves to an adjacent node orstays put. Simultaneously, the rabbit may leap to any node in Zn.
When does the game end?
At “capture time”, when the hunterand the rabbit occupy the samelocation in Zn at the same time.
Goals
Hunter: Minimize “capture time”Rabbit: Maximize “capture time”
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 27: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/27.jpg)
Definition of the game Gn
Rules
At time 0 both hunter and rabbit choose initial positions.
At each subsequent step, the hunter either moves to an adjacent node orstays put. Simultaneously, the rabbit may leap to any node in Zn.
When does the game end?
At “capture time”, when the hunterand the rabbit occupy the samelocation in Zn at the same time.
Goals
Hunter: Minimize “capture time”Rabbit: Maximize “capture time”
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 28: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/28.jpg)
Definition of the game Gn
Rules
At time 0 both hunter and rabbit choose initial positions.
At each subsequent step, the hunter either moves to an adjacent node orstays put. Simultaneously, the rabbit may leap to any node in Zn.
When does the game end?
At “capture time”, when the hunterand the rabbit occupy the samelocation in Zn at the same time.
Goals
Hunter: Minimize “capture time”Rabbit: Maximize “capture time”
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 29: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/29.jpg)
Definition of the game Gn
Rules
At time 0 both hunter and rabbit choose initial positions.
At each subsequent step, the hunter either moves to an adjacent node orstays put. Simultaneously, the rabbit may leap to any node in Zn.
When does the game end?
At “capture time”, when the hunterand the rabbit occupy the samelocation in Zn at the same time.
Goals
Hunter: Minimize “capture time”Rabbit: Maximize “capture time”
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 30: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/30.jpg)
Definition of the game Gn
Rules
At time 0 both hunter and rabbit choose initial positions.
At each subsequent step, the hunter either moves to an adjacent node orstays put. Simultaneously, the rabbit may leap to any node in Zn.
When does the game end?
At “capture time”, when the hunterand the rabbit occupy the samelocation in Zn at the same time.
Goals
Hunter: Minimize “capture time”Rabbit: Maximize “capture time”
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 31: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/31.jpg)
Definition of the game Gn
Rules
At time 0 both hunter and rabbit choose initial positions.
At each subsequent step, the hunter either moves to an adjacent node orstays put. Simultaneously, the rabbit may leap to any node in Zn.
When does the game end?
At “capture time”, when the hunterand the rabbit occupy the samelocation in Zn at the same time.
Goals
Hunter: Minimize “capture time”Rabbit: Maximize “capture time”
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 32: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/32.jpg)
Definition of the game Gn
Rules
At time 0 both hunter and rabbit choose initial positions.
At each subsequent step, the hunter either moves to an adjacent node orstays put. Simultaneously, the rabbit may leap to any node in Zn.
When does the game end?
At “capture time”, when the hunterand the rabbit occupy the samelocation in Zn at the same time.
Goals
Hunter: Minimize “capture time”Rabbit: Maximize “capture time”
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 33: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/33.jpg)
Definition of the game Gn
Rules
At time 0 both hunter and rabbit choose initial positions.
At each subsequent step, the hunter either moves to an adjacent node orstays put. Simultaneously, the rabbit may leap to any node in Zn.
When does the game end?
At “capture time”, when the hunterand the rabbit occupy the samelocation in Zn at the same time.
Goals
Hunter: Minimize “capture time”
Rabbit: Maximize “capture time”
rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 34: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/34.jpg)
Definition of the game Gn
Rules
At time 0 both hunter and rabbit choose initial positions.
At each subsequent step, the hunter either moves to an adjacent node orstays put. Simultaneously, the rabbit may leap to any node in Zn.
When does the game end?
At “capture time”, when the hunterand the rabbit occupy the samelocation in Zn at the same time.
Goals
Hunter: Minimize “capture time”Rabbit: Maximize “capture time”rrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 35: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/35.jpg)
The n-step game G∗n
Define a zero sum game G∗n with payoff 1 to the hunter if he captures therabbit in the first n steps, and payoff 0 otherwise.
G∗n is finite⇒ By the minimax theorem, ∃ optimal randomizedstrategies for both players.
The value of G∗n is the probability pn of capture under optimal play.
Mean capture time in Gn under optimal play is between n/pn and 2n/pn.
We will estimate pn, and construct a Kakeya set of area � pn, thatconsists of 4n triangles.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 36: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/36.jpg)
The n-step game G∗n
Define a zero sum game G∗n with payoff 1 to the hunter if he captures therabbit in the first n steps, and payoff 0 otherwise.
G∗n is finite⇒ By the minimax theorem, ∃ optimal randomizedstrategies for both players.
The value of G∗n is the probability pn of capture under optimal play.
Mean capture time in Gn under optimal play is between n/pn and 2n/pn.
We will estimate pn, and construct a Kakeya set of area � pn, thatconsists of 4n triangles.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 37: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/37.jpg)
The n-step game G∗n
Define a zero sum game G∗n with payoff 1 to the hunter if he captures therabbit in the first n steps, and payoff 0 otherwise.
G∗n is finite⇒ By the minimax theorem, ∃ optimal randomizedstrategies for both players.
The value of G∗n is the probability pn of capture under optimal play.
Mean capture time in Gn under optimal play is between n/pn and 2n/pn.
We will estimate pn, and construct a Kakeya set of area � pn, thatconsists of 4n triangles.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 38: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/38.jpg)
The n-step game G∗n
Define a zero sum game G∗n with payoff 1 to the hunter if he captures therabbit in the first n steps, and payoff 0 otherwise.
G∗n is finite⇒ By the minimax theorem, ∃ optimal randomizedstrategies for both players.
The value of G∗n is the probability pn of capture under optimal play.
Mean capture time in Gn under optimal play is between n/pn and 2n/pn.
We will estimate pn, and construct a Kakeya set of area � pn, thatconsists of 4n triangles.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 39: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/39.jpg)
The n-step game G∗n
Define a zero sum game G∗n with payoff 1 to the hunter if he captures therabbit in the first n steps, and payoff 0 otherwise.
G∗n is finite⇒ By the minimax theorem, ∃ optimal randomizedstrategies for both players.
The value of G∗n is the probability pn of capture under optimal play.
Mean capture time in Gn under optimal play is between n/pn and 2n/pn.
We will estimate pn, and construct a Kakeya set of area � pn, thatconsists of 4n triangles.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 40: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/40.jpg)
Examples of strategies
If the rabbit chooses a random node and stays there, the hunter cansweep the cycle, so expected capture time is ≤ n.
What if the rabbit jumps to a uniform random node in each step?
Then, for any hunter strategy, he will capture the rabbit with probability1/n at each step, so expected capture time is n − 1.
Zig-Zag hunter strategy: He starts in a random direction, thenswitches direction with probability 1/n at each step.
Rabbit counter-strategy: From a random starting node, the rabbitwalks
√n steps to the right, then jumps 2
√n to the left, and repeats.
The probability of capture in n steps is � n−1/2, so mean capture time isn3/2.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 41: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/41.jpg)
Examples of strategies
If the rabbit chooses a random node and stays there, the hunter cansweep the cycle, so expected capture time is ≤ n.
What if the rabbit jumps to a uniform random node in each step?
Then, for any hunter strategy, he will capture the rabbit with probability1/n at each step, so expected capture time is n − 1.
Zig-Zag hunter strategy: He starts in a random direction, thenswitches direction with probability 1/n at each step.
Rabbit counter-strategy: From a random starting node, the rabbitwalks
√n steps to the right, then jumps 2
√n to the left, and repeats.
The probability of capture in n steps is � n−1/2, so mean capture time isn3/2.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 42: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/42.jpg)
Examples of strategies
If the rabbit chooses a random node and stays there, the hunter cansweep the cycle, so expected capture time is ≤ n.
What if the rabbit jumps to a uniform random node in each step?
Then, for any hunter strategy, he will capture the rabbit with probability1/n at each step, so expected capture time is n − 1.
Zig-Zag hunter strategy: He starts in a random direction, thenswitches direction with probability 1/n at each step.
Rabbit counter-strategy: From a random starting node, the rabbitwalks
√n steps to the right, then jumps 2
√n to the left, and repeats.
The probability of capture in n steps is � n−1/2, so mean capture time isn3/2.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 43: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/43.jpg)
Examples of strategies
If the rabbit chooses a random node and stays there, the hunter cansweep the cycle, so expected capture time is ≤ n.
What if the rabbit jumps to a uniform random node in each step?
Then, for any hunter strategy, he will capture the rabbit with probability1/n at each step, so expected capture time is n − 1.
Zig-Zag hunter strategy: He starts in a random direction, thenswitches direction with probability 1/n at each step.
Rabbit counter-strategy: From a random starting node, the rabbitwalks
√n steps to the right, then jumps 2
√n to the left, and repeats.
The probability of capture in n steps is � n−1/2, so mean capture time isn3/2.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 44: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/44.jpg)
Examples of strategies
If the rabbit chooses a random node and stays there, the hunter cansweep the cycle, so expected capture time is ≤ n.
What if the rabbit jumps to a uniform random node in each step?
Then, for any hunter strategy, he will capture the rabbit with probability1/n at each step, so expected capture time is n − 1.
Zig-Zag hunter strategy: He starts in a random direction, thenswitches direction with probability 1/n at each step.
Rabbit counter-strategy: From a random starting node, the rabbitwalks
√n steps to the right, then jumps 2
√n to the left, and repeats.
The probability of capture in n steps is � n−1/2, so mean capture time isn3/2.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 45: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/45.jpg)
Zig-Zag hunter strategytime
space
time
space
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 46: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/46.jpg)
Hunter’s optimal strategy
It turns out the best the hunter can do is start at a random pointand continue at a random speed.
More formally.... Let a,b be independent uniform on [0, 1]. Let the positionof the hunter at time t be
Ht = dan + bte mod n.
What capture time does this yield? Let R` be the position of the rabbit at time` and Kn the number of collisions, i.e.
Kn =n−1∑i=0
1(Ri = Hi ).
Use second moment method – calculate first and second moments of Kn.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 47: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/47.jpg)
Hunter’s optimal strategy
It turns out the best the hunter can do is start at a random pointand continue at a random speed.
More formally.... Let a,b be independent uniform on [0, 1]. Let the positionof the hunter at time t be
Ht = dan + bte mod n.
What capture time does this yield? Let R` be the position of the rabbit at time` and Kn the number of collisions, i.e.
Kn =n−1∑i=0
1(Ri = Hi ).
Use second moment method – calculate first and second moments of Kn.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 48: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/48.jpg)
Hunter’s optimal strategy
It turns out the best the hunter can do is start at a random pointand continue at a random speed.
More formally....
Let a,b be independent uniform on [0, 1]. Let the positionof the hunter at time t be
Ht = dan + bte mod n.
What capture time does this yield? Let R` be the position of the rabbit at time` and Kn the number of collisions, i.e.
Kn =n−1∑i=0
1(Ri = Hi ).
Use second moment method – calculate first and second moments of Kn.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 49: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/49.jpg)
Hunter’s optimal strategy
It turns out the best the hunter can do is start at a random pointand continue at a random speed.
More formally.... Let a,b be independent uniform on [0, 1].
Let the positionof the hunter at time t be
Ht = dan + bte mod n.
What capture time does this yield? Let R` be the position of the rabbit at time` and Kn the number of collisions, i.e.
Kn =n−1∑i=0
1(Ri = Hi ).
Use second moment method – calculate first and second moments of Kn.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 50: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/50.jpg)
Hunter’s optimal strategy
It turns out the best the hunter can do is start at a random pointand continue at a random speed.
More formally.... Let a,b be independent uniform on [0, 1]. Let the positionof the hunter at time t be
Ht = dan + bte mod n.
What capture time does this yield? Let R` be the position of the rabbit at time` and Kn the number of collisions, i.e.
Kn =n−1∑i=0
1(Ri = Hi ).
Use second moment method – calculate first and second moments of Kn.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 51: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/51.jpg)
Hunter’s optimal strategy
It turns out the best the hunter can do is start at a random pointand continue at a random speed.
More formally.... Let a,b be independent uniform on [0, 1]. Let the positionof the hunter at time t be
Ht = dan + bte mod n.
What capture time does this yield?
Let R` be the position of the rabbit at time` and Kn the number of collisions, i.e.
Kn =n−1∑i=0
1(Ri = Hi ).
Use second moment method – calculate first and second moments of Kn.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 52: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/52.jpg)
Hunter’s optimal strategy
It turns out the best the hunter can do is start at a random pointand continue at a random speed.
More formally.... Let a,b be independent uniform on [0, 1]. Let the positionof the hunter at time t be
Ht = dan + bte mod n.
What capture time does this yield? Let R` be the position of the rabbit at time` and Kn the number of collisions
, i.e.
Kn =n−1∑i=0
1(Ri = Hi ).
Use second moment method – calculate first and second moments of Kn.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 53: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/53.jpg)
Hunter’s optimal strategy
It turns out the best the hunter can do is start at a random pointand continue at a random speed.
More formally.... Let a,b be independent uniform on [0, 1]. Let the positionof the hunter at time t be
Ht = dan + bte mod n.
What capture time does this yield? Let R` be the position of the rabbit at time` and Kn the number of collisions, i.e.
Kn =n−1∑i=0
1(Ri = Hi ).
Use second moment method – calculate first and second moments of Kn.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 54: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/54.jpg)
Hunter’s optimal strategy
It turns out the best the hunter can do is start at a random pointand continue at a random speed.
More formally.... Let a,b be independent uniform on [0, 1]. Let the positionof the hunter at time t be
Ht = dan + bte mod n.
What capture time does this yield? Let R` be the position of the rabbit at time` and Kn the number of collisions, i.e.
Kn =n−1∑i=0
1(Ri = Hi ).
Use second moment method – calculate first and second moments of Kn.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 55: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/55.jpg)
Hunter’s optimal strategy
We will show that P(Kn > 0) & 1log n . Recall Kn =
∑n−1i=0 1(Ri = Hi )
hunterrabbithunterrabbithunterrrrrrrrr Ht = dan + bte mod n
E[Kn] =n−1∑i=0
P(Hi = Ri ) = 1
E[K 2
n
]= E[Kn] +
∑i 6=`
P(Hi = Ri ,H` = R`)
Suffices to show E[K 2
n]. log n
Then by Cauchy-Schwartz
P(Kn > 0) ≥ E[Kn]2
E[K 2
n] &
1log n
.
Enough to prove P(Hi = Ri ,Hi+j = Ri+j ) . 1jn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 56: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/56.jpg)
Hunter’s optimal strategy
We will show that P(Kn > 0) & 1log n .
Recall Kn =∑n−1
i=0 1(Ri = Hi )
hunterrabbithunterrabbithunterrrrrrrrr Ht = dan + bte mod n
E[Kn] =n−1∑i=0
P(Hi = Ri ) = 1
E[K 2
n
]= E[Kn] +
∑i 6=`
P(Hi = Ri ,H` = R`)
Suffices to show E[K 2
n]. log n
Then by Cauchy-Schwartz
P(Kn > 0) ≥ E[Kn]2
E[K 2
n] &
1log n
.
Enough to prove P(Hi = Ri ,Hi+j = Ri+j ) . 1jn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 57: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/57.jpg)
Hunter’s optimal strategy
We will show that P(Kn > 0) & 1log n . Recall Kn =
∑n−1i=0 1(Ri = Hi )
hunterrabbithunterrabbithunterrrrrrrrr
Ht = dan + bte mod n
E[Kn] =n−1∑i=0
P(Hi = Ri ) = 1
E[K 2
n
]= E[Kn] +
∑i 6=`
P(Hi = Ri ,H` = R`)
Suffices to show E[K 2
n]. log n
Then by Cauchy-Schwartz
P(Kn > 0) ≥ E[Kn]2
E[K 2
n] &
1log n
.
Enough to prove P(Hi = Ri ,Hi+j = Ri+j ) . 1jn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 58: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/58.jpg)
Hunter’s optimal strategy
We will show that P(Kn > 0) & 1log n . Recall Kn =
∑n−1i=0 1(Ri = Hi )
hunterrabbithunterrabbithunterrrrrrrrr Ht = dan + bte mod n
E[Kn] =n−1∑i=0
P(Hi = Ri ) = 1
E[K 2
n
]= E[Kn] +
∑i 6=`
P(Hi = Ri ,H` = R`)
Suffices to show E[K 2
n]. log n
Then by Cauchy-Schwartz
P(Kn > 0) ≥ E[Kn]2
E[K 2
n] &
1log n
.
Enough to prove P(Hi = Ri ,Hi+j = Ri+j ) . 1jn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 59: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/59.jpg)
Hunter’s optimal strategy
We will show that P(Kn > 0) & 1log n . Recall Kn =
∑n−1i=0 1(Ri = Hi )
hunterrabbithunterrabbithunterrrrrrrrr Ht = dan + bte mod n
E[Kn] =n−1∑i=0
P(Hi = Ri ) = 1
E[K 2
n
]= E[Kn] +
∑i 6=`
P(Hi = Ri ,H` = R`)
Suffices to show E[K 2
n]. log n
Then by Cauchy-Schwartz
P(Kn > 0) ≥ E[Kn]2
E[K 2
n] &
1log n
.
Enough to prove P(Hi = Ri ,Hi+j = Ri+j ) . 1jn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 60: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/60.jpg)
Hunter’s optimal strategy
We will show that P(Kn > 0) & 1log n . Recall Kn =
∑n−1i=0 1(Ri = Hi )
hunterrabbithunterrabbithunterrrrrrrrr Ht = dan + bte mod n
E[Kn] =n−1∑i=0
P(Hi = Ri ) = 1
E[K 2
n
]= E[Kn] +
∑i 6=`
P(Hi = Ri ,H` = R`)
Suffices to show E[K 2
n]. log n
Then by Cauchy-Schwartz
P(Kn > 0) ≥ E[Kn]2
E[K 2
n] &
1log n
.
Enough to prove P(Hi = Ri ,Hi+j = Ri+j ) . 1jn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 61: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/61.jpg)
Hunter’s optimal strategy
We will show that P(Kn > 0) & 1log n . Recall Kn =
∑n−1i=0 1(Ri = Hi )
hunterrabbithunterrabbithunterrrrrrrrr Ht = dan + bte mod n
E[Kn] =n−1∑i=0
P(Hi = Ri ) = 1
E[K 2
n
]= E[Kn] +
∑i 6=`
P(Hi = Ri ,H` = R`)
Suffices to show E[K 2
n]. log n
Then by Cauchy-Schwartz
P(Kn > 0) ≥ E[Kn]2
E[K 2
n] &
1log n
.
Enough to prove P(Hi = Ri ,Hi+j = Ri+j ) . 1jn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 62: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/62.jpg)
Hunter’s optimal strategy
We will show that P(Kn > 0) & 1log n . Recall Kn =
∑n−1i=0 1(Ri = Hi )
hunterrabbithunterrabbithunterrrrrrrrr Ht = dan + bte mod n
E[Kn] =n−1∑i=0
P(Hi = Ri ) = 1
E[K 2
n
]= E[Kn] +
∑i 6=`
P(Hi = Ri ,H` = R`)
Suffices to show E[K 2
n]. log n
Then by Cauchy-Schwartz
P(Kn > 0) ≥ E[Kn]2
E[K 2
n] &
1log n
.
Enough to prove P(Hi = Ri ,Hi+j = Ri+j ) . 1jn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 63: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/63.jpg)
Hunter’s optimal strategy
We will show that P(Kn > 0) & 1log n . Recall Kn =
∑n−1i=0 1(Ri = Hi )
hunterrabbithunterrabbithunterrrrrrrrr Ht = dan + bte mod n
E[Kn] =n−1∑i=0
P(Hi = Ri ) = 1
E[K 2
n
]= E[Kn] +
∑i 6=`
P(Hi = Ri ,H` = R`)
Suffices to show E[K 2
n]. log n
Then by Cauchy-Schwartz
P(Kn > 0) ≥ E[Kn]2
E[K 2
n] &
1log n
.
Enough to prove P(Hi = Ri ,Hi+j = Ri+j ) . 1jn
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 64: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/64.jpg)
Hunter’s optimal strategy
Need to prove
P(Hi = Ri ,Hi+j = Ri+j ) .1jn.
This is equivalent to showing that for r , s fixed Recall a, b ∼ U[0, 1]
P(an + bi ∈ (r − 1, r ], na + b(i + j) ∈ (s − 1, s]) .1jn.
Subtract the two constraints to get bj ∈ [s − r − 1, s − r + 1] – this hasmeasure at most 2/j .
After fixing b, the choices for a have measure 1/n.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 65: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/65.jpg)
Hunter’s optimal strategy
Need to prove
P(Hi = Ri ,Hi+j = Ri+j ) .1jn.
This is equivalent to showing that for r , s fixed Recall a, b ∼ U[0, 1]
P(an + bi ∈ (r − 1, r ], na + b(i + j) ∈ (s − 1, s]) .1jn.
Subtract the two constraints to get bj ∈ [s − r − 1, s − r + 1] – this hasmeasure at most 2/j .
After fixing b, the choices for a have measure 1/n.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 66: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/66.jpg)
Hunter’s optimal strategy
Need to prove
P(Hi = Ri ,Hi+j = Ri+j ) .1jn.
This is equivalent to showing that for r , s fixed
Recall a, b ∼ U[0, 1]
P(an + bi ∈ (r − 1, r ], na + b(i + j) ∈ (s − 1, s]) .1jn.
Subtract the two constraints to get bj ∈ [s − r − 1, s − r + 1] – this hasmeasure at most 2/j .
After fixing b, the choices for a have measure 1/n.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 67: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/67.jpg)
Hunter’s optimal strategy
Need to prove
P(Hi = Ri ,Hi+j = Ri+j ) .1jn.
This is equivalent to showing that for r , s fixed Recall a, b ∼ U[0, 1]
P(an + bi ∈ (r − 1, r ], na + b(i + j) ∈ (s − 1, s]) .1jn.
Subtract the two constraints to get bj ∈ [s − r − 1, s − r + 1] – this hasmeasure at most 2/j .
After fixing b, the choices for a have measure 1/n.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 68: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/68.jpg)
Hunter’s optimal strategy
Need to prove
P(Hi = Ri ,Hi+j = Ri+j ) .1jn.
This is equivalent to showing that for r , s fixed Recall a, b ∼ U[0, 1]
P(an + bi ∈ (r − 1, r ], na + b(i + j) ∈ (s − 1, s]) .1jn.
Subtract the two constraints to get bj ∈ [s − r − 1, s − r + 1] – this hasmeasure at most 2/j .
After fixing b, the choices for a have measure 1/n.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 69: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/69.jpg)
Hunter’s optimal strategy
Need to prove
P(Hi = Ri ,Hi+j = Ri+j ) .1jn.
This is equivalent to showing that for r , s fixed Recall a, b ∼ U[0, 1]
P(an + bi ∈ (r − 1, r ], na + b(i + j) ∈ (s − 1, s]) .1jn.
Subtract the two constraints to get bj ∈ [s − r − 1, s − r + 1] – this hasmeasure at most 2/j .
After fixing b, the choices for a have measure 1/n.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 70: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/70.jpg)
Hunter’s optimal strategy
Need to prove
P(Hi = Ri ,Hi+j = Ri+j ) .1jn.
This is equivalent to showing that for r , s fixed Recall a, b ∼ U[0, 1]
P(an + bi ∈ (r − 1, r ], na + b(i + j) ∈ (s − 1, s]) .1jn.
Subtract the two constraints to get bj ∈ [s − r − 1, s − r + 1] – this hasmeasure at most 2/j .
After fixing b, the choices for a have measure 1/n.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 71: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/71.jpg)
Hunter’s optimal strategy
Need to prove
P(Hi = Ri ,Hi+j = Ri+j ) .1jn.
This is equivalent to showing that for r , s fixed Recall a, b ∼ U[0, 1]
P(an + bi ∈ (r − 1, r ], na + b(i + j) ∈ (s − 1, s]) .1jn.
Subtract the two constraints to get bj ∈ [s − r − 1, s − r + 1] – this hasmeasure at most 2/j .
After fixing b, the choices for a have measure 1/n.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 72: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/72.jpg)
Rabbit’s optimal strategy
With the hunter’s strategy above rrr Recall Kn =∑n−1
i=0 1(Hi = Ri )
P(Kn > 0) &1
log n.
This gave expected capture time at most n log n.What about the rabbit? Can he escape for time of order n log n?Looking for a rabbit strategy with
P(Kn > 0) .1
log n.
Extend the strategies until time 2n and define K2n analogously. Obviously
P(Kn > 0) ≤ E[K2n]
E[K2n | Kn > 0]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 73: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/73.jpg)
Rabbit’s optimal strategy
With the hunter’s strategy above rrr
Recall Kn =∑n−1
i=0 1(Hi = Ri )
P(Kn > 0) &1
log n.
This gave expected capture time at most n log n.What about the rabbit? Can he escape for time of order n log n?Looking for a rabbit strategy with
P(Kn > 0) .1
log n.
Extend the strategies until time 2n and define K2n analogously. Obviously
P(Kn > 0) ≤ E[K2n]
E[K2n | Kn > 0]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 74: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/74.jpg)
Rabbit’s optimal strategy
With the hunter’s strategy above rrr Recall Kn =∑n−1
i=0 1(Hi = Ri )
P(Kn > 0) &1
log n.
This gave expected capture time at most n log n.What about the rabbit? Can he escape for time of order n log n?Looking for a rabbit strategy with
P(Kn > 0) .1
log n.
Extend the strategies until time 2n and define K2n analogously. Obviously
P(Kn > 0) ≤ E[K2n]
E[K2n | Kn > 0]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 75: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/75.jpg)
Rabbit’s optimal strategy
With the hunter’s strategy above rrr Recall Kn =∑n−1
i=0 1(Hi = Ri )
P(Kn > 0) &1
log n.
This gave expected capture time at most n log n.What about the rabbit? Can he escape for time of order n log n?Looking for a rabbit strategy with
P(Kn > 0) .1
log n.
Extend the strategies until time 2n and define K2n analogously. Obviously
P(Kn > 0) ≤ E[K2n]
E[K2n | Kn > 0]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 76: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/76.jpg)
Rabbit’s optimal strategy
With the hunter’s strategy above rrr Recall Kn =∑n−1
i=0 1(Hi = Ri )
P(Kn > 0) &1
log n.
This gave expected capture time at most n log n.
What about the rabbit? Can he escape for time of order n log n?Looking for a rabbit strategy with
P(Kn > 0) .1
log n.
Extend the strategies until time 2n and define K2n analogously. Obviously
P(Kn > 0) ≤ E[K2n]
E[K2n | Kn > 0]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 77: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/77.jpg)
Rabbit’s optimal strategy
With the hunter’s strategy above rrr Recall Kn =∑n−1
i=0 1(Hi = Ri )
P(Kn > 0) &1
log n.
This gave expected capture time at most n log n.What about the rabbit?
Can he escape for time of order n log n?Looking for a rabbit strategy with
P(Kn > 0) .1
log n.
Extend the strategies until time 2n and define K2n analogously. Obviously
P(Kn > 0) ≤ E[K2n]
E[K2n | Kn > 0]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 78: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/78.jpg)
Rabbit’s optimal strategy
With the hunter’s strategy above rrr Recall Kn =∑n−1
i=0 1(Hi = Ri )
P(Kn > 0) &1
log n.
This gave expected capture time at most n log n.What about the rabbit? Can he escape for time of order n log n?
Looking for a rabbit strategy with
P(Kn > 0) .1
log n.
Extend the strategies until time 2n and define K2n analogously. Obviously
P(Kn > 0) ≤ E[K2n]
E[K2n | Kn > 0]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 79: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/79.jpg)
Rabbit’s optimal strategy
With the hunter’s strategy above rrr Recall Kn =∑n−1
i=0 1(Hi = Ri )
P(Kn > 0) &1
log n.
This gave expected capture time at most n log n.What about the rabbit? Can he escape for time of order n log n?Looking for a rabbit strategy with
P(Kn > 0) .1
log n.
Extend the strategies until time 2n and define K2n analogously. Obviously
P(Kn > 0) ≤ E[K2n]
E[K2n | Kn > 0]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 80: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/80.jpg)
Rabbit’s optimal strategy
With the hunter’s strategy above rrr Recall Kn =∑n−1
i=0 1(Hi = Ri )
P(Kn > 0) &1
log n.
This gave expected capture time at most n log n.What about the rabbit? Can he escape for time of order n log n?Looking for a rabbit strategy with
P(Kn > 0) .1
log n.
Extend the strategies until time 2n and define K2n analogously.
Obviously
P(Kn > 0) ≤ E[K2n]
E[K2n | Kn > 0]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 81: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/81.jpg)
Rabbit’s optimal strategy
With the hunter’s strategy above rrr Recall Kn =∑n−1
i=0 1(Hi = Ri )
P(Kn > 0) &1
log n.
This gave expected capture time at most n log n.What about the rabbit? Can he escape for time of order n log n?Looking for a rabbit strategy with
P(Kn > 0) .1
log n.
Extend the strategies until time 2n and define K2n analogously. Obviously
P(Kn > 0) ≤ E[K2n]
E[K2n | Kn > 0]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 82: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/82.jpg)
Rabbit’s optimal strategy
If the rabbit starts at a uniform point and the jumps are independent, then
E[K2n] = 2 rrrrrr Recall K2n =2n−1∑i=0
1(Hi = Ri )
Idea: Need to make E[K2n | Kn > 0] “big” so P(Kn > 0) ≤ (log n)−1.
This means that given the rabbit and hunter collided, we want them tocollide “a lot”. The hunter can only move to neighbours or stay put.
So the rabbit should also choose a distribution for the jumps that favors shortdistances, yet grows linearly in time. This suggests a Cauchy random walk.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 83: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/83.jpg)
Rabbit’s optimal strategy
If the rabbit starts at a uniform point and the jumps are independent, then
E[K2n] = 2 rrrrrr Recall K2n =2n−1∑i=0
1(Hi = Ri )
Idea: Need to make E[K2n | Kn > 0] “big” so P(Kn > 0) ≤ (log n)−1.
This means that given the rabbit and hunter collided, we want them tocollide “a lot”. The hunter can only move to neighbours or stay put.
So the rabbit should also choose a distribution for the jumps that favors shortdistances, yet grows linearly in time. This suggests a Cauchy random walk.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 84: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/84.jpg)
Rabbit’s optimal strategy
If the rabbit starts at a uniform point and the jumps are independent, then
E[K2n] = 2 rrrrrr Recall K2n =2n−1∑i=0
1(Hi = Ri )
Idea: Need to make E[K2n | Kn > 0] “big” so P(Kn > 0) ≤ (log n)−1.
This means that given the rabbit and hunter collided, we want them tocollide “a lot”. The hunter can only move to neighbours or stay put.
So the rabbit should also choose a distribution for the jumps that favors shortdistances, yet grows linearly in time. This suggests a Cauchy random walk.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 85: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/85.jpg)
Rabbit’s optimal strategy
If the rabbit starts at a uniform point and the jumps are independent, then
E[K2n] = 2 rrrrrr Recall K2n =2n−1∑i=0
1(Hi = Ri )
Idea: Need to make E[K2n | Kn > 0] “big” so P(Kn > 0) ≤ (log n)−1.
This means that given the rabbit and hunter collided, we want them tocollide “a lot”. The hunter can only move to neighbours or stay put.
So the rabbit should also choose a distribution for the jumps that favors shortdistances, yet grows linearly in time. This suggests a Cauchy random walk.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 86: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/86.jpg)
Rabbit’s optimal strategy
If the rabbit starts at a uniform point and the jumps are independent, then
E[K2n] = 2 rrrrrr Recall K2n =2n−1∑i=0
1(Hi = Ri )
Idea: Need to make E[K2n | Kn > 0] “big” so P(Kn > 0) ≤ (log n)−1.
This means that given the rabbit and hunter collided, we want them tocollide “a lot”. The hunter can only move to neighbours or stay put.
So the rabbit should also choose a distribution for the jumps that favors shortdistances, yet grows linearly in time. This suggests a Cauchy random walk.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 87: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/87.jpg)
Rabbit’s optimal strategy
If the rabbit starts at a uniform point and the jumps are independent, then
E[K2n] = 2 rrrrrr Recall K2n =2n−1∑i=0
1(Hi = Ri )
Idea: Need to make E[K2n | Kn > 0] “big” so P(Kn > 0) ≤ (log n)−1.
This means that given the rabbit and hunter collided, we want them tocollide “a lot”. The hunter can only move to neighbours or stay put.
So the rabbit should also choose a distribution for the jumps that favors shortdistances, yet grows linearly in time. This suggests a Cauchy random walk.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 88: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/88.jpg)
Cauchy Rabbit
By time i the hunter can only be in the set {−i mod n, . . . , i mod n}. We arelooking for a distribution for the rabbit so that
P(Ri = `) &1i
for ` ∈ {−i mod n, . . . , i mod n}.
Then by the Markov property
E[K2n | Kn > 0] ≥n−1∑i=0
P0(Hi = Ri ) & log n.
Intuition: If X1, . . . are i.i.d. Cauchy random variables, i.e. with density
(π(1 + x2))−1, then X1 + . . .+ Xn is spread over (−n, n) and with roughly
uniform distribution.
This is what we want- But in the discrete setting...
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 89: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/89.jpg)
Cauchy Rabbit
By time i the hunter can only be in the set {−i mod n, . . . , i mod n}. We arelooking for a distribution for the rabbit so that
P(Ri = `) &1i
for ` ∈ {−i mod n, . . . , i mod n}.
Then by the Markov property
E[K2n | Kn > 0] ≥n−1∑i=0
P0(Hi = Ri ) & log n.
Intuition: If X1, . . . are i.i.d. Cauchy random variables, i.e. with density
(π(1 + x2))−1, then X1 + . . .+ Xn is spread over (−n, n) and with roughly
uniform distribution.
This is what we want- But in the discrete setting...
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 90: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/90.jpg)
Cauchy Rabbit
By time i the hunter can only be in the set {−i mod n, . . . , i mod n}. We arelooking for a distribution for the rabbit so that
P(Ri = `) &1i
for ` ∈ {−i mod n, . . . , i mod n}.
Then by the Markov property
E[K2n | Kn > 0] ≥n−1∑i=0
P0(Hi = Ri ) & log n.
Intuition: If X1, . . . are i.i.d. Cauchy random variables, i.e. with density
(π(1 + x2))−1, then X1 + . . .+ Xn is spread over (−n, n) and with roughly
uniform distribution.
This is what we want- But in the discrete setting...
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 91: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/91.jpg)
Cauchy Rabbit
By time i the hunter can only be in the set {−i mod n, . . . , i mod n}. We arelooking for a distribution for the rabbit so that
P(Ri = `) &1i
for ` ∈ {−i mod n, . . . , i mod n}.
Then by the Markov property
E[K2n | Kn > 0] ≥n−1∑i=0
P0(Hi = Ri ) & log n.
Intuition: If X1, . . . are i.i.d. Cauchy random variables, i.e. with density
(π(1 + x2))−1, then X1 + . . .+ Xn is spread over (−n, n) and with roughly
uniform distribution.
This is what we want- But in the discrete setting...
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 92: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/92.jpg)
Cauchy Rabbit
By time i the hunter can only be in the set {−i mod n, . . . , i mod n}. We arelooking for a distribution for the rabbit so that
P(Ri = `) &1i
for ` ∈ {−i mod n, . . . , i mod n}.
Then by the Markov property
E[K2n | Kn > 0] ≥n−1∑i=0
P0(Hi = Ri ) & log n.
Intuition: If X1, . . . are i.i.d. Cauchy random variables, i.e. with density
(π(1 + x2))−1, then X1 + . . .+ Xn is spread over (−n, n) and with roughly
uniform distribution.
This is what we want- But in the discrete setting...
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 93: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/93.jpg)
Cauchy Rabbit
By time i the hunter can only be in the set {−i mod n, . . . , i mod n}. We arelooking for a distribution for the rabbit so that
P(Ri = `) &1i
for ` ∈ {−i mod n, . . . , i mod n}.
Then by the Markov property
E[K2n | Kn > 0] ≥n−1∑i=0
P0(Hi = Ri ) & log n.
Intuition: If X1, . . . are i.i.d. Cauchy random variables, i.e. with density
(π(1 + x2))−1, then X1 + . . .+ Xn is spread over (−n, n) and with roughly
uniform distribution.
This is what we want- But in the discrete setting...
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 94: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/94.jpg)
Cauchy RabbitThe Cauchy distribution can be embedded in planar Brownian motion.
Let’s imitate that in the discrete setting:Let (Xt ,Yt )t be a simple random walk in Z2.
Define hitting times
Ti = inf{t ≥ 0 : Yt = i}
and set Ri = XTi modn.
With probability 1/4, SRW exitsthe square via the top side.
Of the 2i + 1 nodes on the top,the middle node is the mostlikely hitting point: subdivide alledges, and condition on the(even) number of horizontalsteps until height i is reached;the horizontal displacement isa shifted binomial, so the modeis the mean.
Thus the hitting probability at(0, i) is at least 1/(8i + 4).
)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 95: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/95.jpg)
Cauchy RabbitThe Cauchy distribution can be embedded in planar Brownian motion.
Let’s imitate that in the discrete setting:
Let (Xt ,Yt )t be a simple random walk in Z2.Define hitting times
Ti = inf{t ≥ 0 : Yt = i}
and set Ri = XTi modn.
With probability 1/4, SRW exitsthe square via the top side.
Of the 2i + 1 nodes on the top,the middle node is the mostlikely hitting point: subdivide alledges, and condition on the(even) number of horizontalsteps until height i is reached;the horizontal displacement isa shifted binomial, so the modeis the mean.
Thus the hitting probability at(0, i) is at least 1/(8i + 4).
)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 96: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/96.jpg)
Cauchy RabbitThe Cauchy distribution can be embedded in planar Brownian motion.
Let’s imitate that in the discrete setting:Let (Xt ,Yt )t be a simple random walk in Z2.Define hitting times
Ti = inf{t ≥ 0 : Yt = i}
and set Ri = XTi modn.
With probability 1/4, SRW exitsthe square via the top side.
Of the 2i + 1 nodes on the top,the middle node is the mostlikely hitting point: subdivide alledges, and condition on the(even) number of horizontalsteps until height i is reached;the horizontal displacement isa shifted binomial, so the modeis the mean.
Thus the hitting probability at(0, i) is at least 1/(8i + 4).
)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 97: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/97.jpg)
Cauchy RabbitThe Cauchy distribution can be embedded in planar Brownian motion.
Let’s imitate that in the discrete setting:Let (Xt ,Yt )t be a simple random walk in Z2.Define hitting times
Ti = inf{t ≥ 0 : Yt = i}
and set Ri = XTi modn.
With probability 1/4, SRW exitsthe square via the top side.
Of the 2i + 1 nodes on the top,the middle node is the mostlikely hitting point: subdivide alledges, and condition on the(even) number of horizontalsteps until height i is reached;the horizontal displacement isa shifted binomial, so the modeis the mean.
Thus the hitting probability at(0, i) is at least 1/(8i + 4).
)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 98: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/98.jpg)
Cauchy RabbitThe Cauchy distribution can be embedded in planar Brownian motion.
Let’s imitate that in the discrete setting:Let (Xt ,Yt )t be a simple random walk in Z2.Define hitting times
Ti = inf{t ≥ 0 : Yt = i}
and set Ri = XTi modn.
With probability 1/4, SRW exitsthe square via the top side.
Of the 2i + 1 nodes on the top,the middle node is the mostlikely hitting point: subdivide alledges, and condition on the(even) number of horizontalsteps until height i is reached;the horizontal displacement isa shifted binomial, so the modeis the mean.
Thus the hitting probability at(0, i) is at least 1/(8i + 4).
)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 99: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/99.jpg)
Cauchy RabbitThe Cauchy distribution can be embedded in planar Brownian motion.
Let’s imitate that in the discrete setting:Let (Xt ,Yt )t be a simple random walk in Z2.Define hitting times
Ti = inf{t ≥ 0 : Yt = i}
and set Ri = XTi modn.
With probability 1/4, SRW exitsthe square via the top side.
Of the 2i + 1 nodes on the top,the middle node is the mostlikely hitting point: subdivide alledges, and condition on the(even) number of horizontalsteps until height i is reached;the horizontal displacement isa shifted binomial, so the modeis the mean.
Thus the hitting probability at(0, i) is at least 1/(8i + 4).
)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 100: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/100.jpg)
Cauchy RabbitThe Cauchy distribution can be embedded in planar Brownian motion.
Let’s imitate that in the discrete setting:Let (Xt ,Yt )t be a simple random walk in Z2.Define hitting times
Ti = inf{t ≥ 0 : Yt = i}
and set Ri = XTi modn.
With probability 1/4, SRW exitsthe square via the top side.
Of the 2i + 1 nodes on the top,the middle node is the mostlikely hitting point: subdivide alledges, and condition on the(even) number of horizontalsteps until height i is reached;the horizontal displacement isa shifted binomial, so the modeis the mean.
Thus the hitting probability at(0, i) is at least 1/(8i + 4).
)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 101: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/101.jpg)
Cauchy Rabbit
Suppose 0 < k < i .With probability 1/4, SRW exits thesquare [−k , k ]2 via the right side.Repeating the previous argument,the hitting probability at (k , i) is atleast c/i .
time
space
(k,k)
(k,ïk)
(ïk,k)
(ïk,ïk)
(0,0)
(k,i)
(k,l)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 102: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/102.jpg)
Cauchy Rabbit
Suppose 0 < k < i .
With probability 1/4, SRW exits thesquare [−k , k ]2 via the right side.Repeating the previous argument,the hitting probability at (k , i) is atleast c/i .
time
space
(k,k)
(k,ïk)
(ïk,k)
(ïk,ïk)
(0,0)
(k,i)
(k,l)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 103: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/103.jpg)
Cauchy Rabbit
Suppose 0 < k < i .With probability 1/4, SRW exits thesquare [−k , k ]2 via the right side.
Repeating the previous argument,the hitting probability at (k , i) is atleast c/i .
time
space
(k,k)
(k,ïk)
(ïk,k)
(ïk,ïk)
(0,0)
(k,i)
(k,l)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 104: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/104.jpg)
Cauchy Rabbit
Suppose 0 < k < i .With probability 1/4, SRW exits thesquare [−k , k ]2 via the right side.Repeating the previous argument,the hitting probability at (k , i) is atleast c/i .
time
space
(k,k)
(k,ïk)
(ïk,k)
(ïk,ïk)
(0,0)
(k,i)
(k,l)
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 105: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/105.jpg)
Rabbit and Hunter construct Kakeya sets
Let (Rt )t be a rabbit strategy. Extend it to real times as a step function.
Let a be uniform in [−1, 1] and b uniform in [0, 1] and Ht = an + bt . There isa collision at time t ∈ [0, n) if Rt = Ht .
What is the chance there is a collision in [m,m + 1)?
It is P(an + bm ≤ Rm < an + b(m + 1)), which is half the area of the triangle
Rm
na
b
1
1
−1
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 106: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/106.jpg)
Rabbit and Hunter construct Kakeya sets
Let (Rt )t be a rabbit strategy. Extend it to real times as a step function.Let a be uniform in [−1, 1] and b uniform in [0, 1] and Ht = an + bt . There isa collision at time t ∈ [0, n) if Rt = Ht .
What is the chance there is a collision in [m,m + 1)?
It is P(an + bm ≤ Rm < an + b(m + 1)), which is half the area of the triangle
Rm
na
b
1
1
−1
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 107: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/107.jpg)
Rabbit and Hunter construct Kakeya sets
Let (Rt )t be a rabbit strategy. Extend it to real times as a step function.Let a be uniform in [−1, 1] and b uniform in [0, 1] and Ht = an + bt . There isa collision at time t ∈ [0, n) if Rt = Ht .
What is the chance there is a collision in [m,m + 1)?
It is P(an + bm ≤ Rm < an + b(m + 1)), which is half the area of the triangle
Rm
na
b
1
1
−1
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 108: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/108.jpg)
Rabbit and Hunter construct Kakeya sets
Let (Rt )t be a rabbit strategy. Extend it to real times as a step function.Let a be uniform in [−1, 1] and b uniform in [0, 1] and Ht = an + bt . There isa collision at time t ∈ [0, n) if Rt = Ht .
What is the chance there is a collision in [m,m + 1)?
It is P(an + bm ≤ Rm < an + b(m + 1)), which is half the area of the triangle
Rm
na
b
1
1
−1
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 109: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/109.jpg)
Rabbit and Hunter construct Kakeya sets
Let (Rt )t be a rabbit strategy. Extend it to real times as a step function.Let a be uniform in [−1, 1] and b uniform in [0, 1] and Ht = an + bt . There isa collision at time t ∈ [0, n) if Rt = Ht .
What is the chance there is a collision in [m,m + 1)?
It is P(an + bm ≤ Rm < an + b(m + 1)), which is half the area of the triangle
Rm
na
b
1
1
−1
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 110: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/110.jpg)
Rabbit and Hunter construct Kakeya sets
Let (Rt )t be a rabbit strategy. Extend it to real times as a step function.Let a be uniform in [−1, 1] and b uniform in [0, 1] and Ht = an + bt . There isa collision at time t ∈ [0, n) if Rt = Ht .
What is the chance there is a collision in [m,m + 1)?
It is P(an + bm ≤ Rm < an + b(m + 1)), which is half the area of the triangle
Rm
na
b
1
1
−1
an+ bm = Rman+ b(m+ 1) = Rm
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 111: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/111.jpg)
Rabbit and Hunter construct Kakeya sets
Hence the probability of collision in [0, n) is half the area of the union of allsuch triangles, which are translates of
T1T2Tn Tn−1
1
1n
1n
. . .
In these triangles we can find a unit segment in all directions that have anangle in [0, π/4]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 112: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/112.jpg)
Rabbit and Hunter construct Kakeya sets
Hence the probability of collision in [0, n) is half the area of the union of allsuch triangles, which are translates of
T1T2Tn Tn−1
1
1n
1n
. . .
In these triangles we can find a unit segment in all directions that have anangle in [0, π/4]
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 113: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/113.jpg)
Rabbit and Hunter construct Kakeya setsIf the rabbit employs the Cauchy strategy, then
P(collision in the first n steps) .1
log n.
Hence, this gives a set of triangles with area of order at most 1/ log n.
Simulation generated with n = 32
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 114: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/114.jpg)
Rabbit and Hunter construct Kakeya setsIf the rabbit employs the Cauchy strategy, then
P(collision in the first n steps) .1
log n.
Hence, this gives a set of triangles with area of order at most 1/ log n.
Simulation generated with n = 32
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 115: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/115.jpg)
Rabbit and Hunter construct Kakeya setsIf the rabbit employs the Cauchy strategy, then
P(collision in the first n steps) .1
log n.
Hence, this gives a set of triangles with area of order at most 1/ log n.
Simulation generated with n = 32Yuval Peres Search Games and Optimal Kakeya Sets
![Page 116: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/116.jpg)
The Cauchy process {Xt} can be embedded in planarBrownian motion.
(Xt, t)
t
Xt+s − Xt has the same law as tX1 and X1 has the Cauchy distribution(density given by (π(1 + x2))−1).
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 117: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/117.jpg)
The Cauchy process {Xt} can be embedded in planarBrownian motion.
(Xt, t)
t
Xt+s − Xt has the same law as tX1 and X1 has the Cauchy distribution(density given by (π(1 + x2))−1).
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 118: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/118.jpg)
The Cauchy process {Xt} can be embedded in planarBrownian motion.
(Xt, t)
t
Xt+s − Xt has the same law as tX1 and X1 has the Cauchy distribution(density given by (π(1 + x2))−1).
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 119: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/119.jpg)
Kakeya sets from the Cauchy process
Motivated by the Cauchy strategy, let’s see a continuum analog of theprobabilistic Kakeya construction of the hunter and rabbit.
Let (Xt )t be a Cauchy process. Set
Λ = {(a,Xt + at) : a, t ∈ [0, 1]}.
Λ is a quarter of a Kakeya set – it contains all directions from 0 up to 45◦
degrees. Take four rotated copies of Λ to obtain a Kakeya set.
Λ is an optimal Kakeya set!
Leb(Λ) = 0 and most importantly the ε-neighbourhood satisfies almostsurely
Leb(Λ(ε)) � 1| log ε|
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 120: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/120.jpg)
Kakeya sets from the Cauchy process
Motivated by the Cauchy strategy, let’s see a continuum analog of theprobabilistic Kakeya construction of the hunter and rabbit.
Let (Xt )t be a Cauchy process. Set
Λ = {(a,Xt + at) : a, t ∈ [0, 1]}.
Λ is a quarter of a Kakeya set – it contains all directions from 0 up to 45◦
degrees. Take four rotated copies of Λ to obtain a Kakeya set.
Λ is an optimal Kakeya set!
Leb(Λ) = 0 and most importantly the ε-neighbourhood satisfies almostsurely
Leb(Λ(ε)) � 1| log ε|
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 121: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/121.jpg)
Kakeya sets from the Cauchy process
Motivated by the Cauchy strategy, let’s see a continuum analog of theprobabilistic Kakeya construction of the hunter and rabbit.
Let (Xt )t be a Cauchy process. Set
Λ = {(a,Xt + at) : a, t ∈ [0, 1]}.
Λ is a quarter of a Kakeya set – it contains all directions from 0 up to 45◦
degrees. Take four rotated copies of Λ to obtain a Kakeya set.
Λ is an optimal Kakeya set!
Leb(Λ) = 0 and most importantly the ε-neighbourhood satisfies almostsurely
Leb(Λ(ε)) � 1| log ε|
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 122: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/122.jpg)
Kakeya sets from the Cauchy process
Motivated by the Cauchy strategy, let’s see a continuum analog of theprobabilistic Kakeya construction of the hunter and rabbit.
Let (Xt )t be a Cauchy process. Set
Λ = {(a,Xt + at) : a, t ∈ [0, 1]}.
Λ is a quarter of a Kakeya set – it contains all directions from 0 up to 45◦
degrees. Take four rotated copies of Λ to obtain a Kakeya set.
Λ is an optimal Kakeya set!
Leb(Λ) = 0 and most importantly the ε-neighbourhood satisfies almostsurely
Leb(Λ(ε)) � 1| log ε|
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 123: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/123.jpg)
Kakeya sets from the Cauchy process
Motivated by the Cauchy strategy, let’s see a continuum analog of theprobabilistic Kakeya construction of the hunter and rabbit.
Let (Xt )t be a Cauchy process. Set
Λ = {(a,Xt + at) : a, t ∈ [0, 1]}.
Λ is a quarter of a Kakeya set – it contains all directions from 0 up to 45◦
degrees. Take four rotated copies of Λ to obtain a Kakeya set.
Λ is an optimal Kakeya set!
Leb(Λ) = 0 and most importantly the ε-neighbourhood satisfies almostsurely
Leb(Λ(ε)) � 1| log ε|
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 124: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/124.jpg)
Kakeya sets – Open problems
Keich in 1999 showed there is no Kakeya set which is a union of n triangleswith area of smaller order than 1/ log n. Bourgain earlier noted that the εneighborhood of any Kakeya set has area at least 1/|logε|.
So the random construction is optimal.
Davies in 1971 showed that Kakeya sets in the plane have Hausdorffdimension equal to 2.
It is a major open problem whether Kakeya sets in dimensions d > 2 haveHausdorff dimension equal to d .
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 125: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/125.jpg)
Kakeya sets – Open problems
Keich in 1999 showed there is no Kakeya set which is a union of n triangleswith area of smaller order than 1/ log n. Bourgain earlier noted that the εneighborhood of any Kakeya set has area at least 1/|logε|.
So the random construction is optimal.
Davies in 1971 showed that Kakeya sets in the plane have Hausdorffdimension equal to 2.
It is a major open problem whether Kakeya sets in dimensions d > 2 haveHausdorff dimension equal to d .
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 126: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/126.jpg)
Kakeya sets – Open problems
Keich in 1999 showed there is no Kakeya set which is a union of n triangleswith area of smaller order than 1/ log n. Bourgain earlier noted that the εneighborhood of any Kakeya set has area at least 1/|logε|.
So the random construction is optimal.
Davies in 1971 showed that Kakeya sets in the plane have Hausdorffdimension equal to 2.
It is a major open problem whether Kakeya sets in dimensions d > 2 haveHausdorff dimension equal to d .
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 127: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/127.jpg)
Kakeya sets – Open problems
Keich in 1999 showed there is no Kakeya set which is a union of n triangleswith area of smaller order than 1/ log n. Bourgain earlier noted that the εneighborhood of any Kakeya set has area at least 1/|logε|.
So the random construction is optimal.
Davies in 1971 showed that Kakeya sets in the plane have Hausdorffdimension equal to 2.
It is a major open problem whether Kakeya sets in dimensions d > 2 haveHausdorff dimension equal to d .
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 128: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/128.jpg)
Kakeya sets – Open problems
Keich in 1999 showed there is no Kakeya set which is a union of n triangleswith area of smaller order than 1/ log n. Bourgain earlier noted that the εneighborhood of any Kakeya set has area at least 1/|logε|.
So the random construction is optimal.
Davies in 1971 showed that Kakeya sets in the plane have Hausdorffdimension equal to 2.
It is a major open problem whether Kakeya sets in dimensions d > 2 haveHausdorff dimension equal to d .
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 129: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/129.jpg)
General graphs
Consider a graph onn vertices.
Pick a spanning tree.
Depth first search yields
This is a closed path oflength 2n − 2.
The hunter can nowemploy his previousstrategy on this path.This will give O(n log n)capture time.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 130: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/130.jpg)
General graphs
Consider a graph onn vertices.
Pick a spanning tree.
Depth first search yields
This is a closed path oflength 2n − 2.
The hunter can nowemploy his previousstrategy on this path.This will give O(n log n)capture time.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 131: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/131.jpg)
General graphs
Consider a graph onn vertices.
Pick a spanning tree.
Depth first search yields
This is a closed path oflength 2n − 2.
The hunter can nowemploy his previousstrategy on this path.This will give O(n log n)capture time.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 132: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/132.jpg)
General graphs
Consider a graph onn vertices.
Pick a spanning tree.
Depth first search yields
This is a closed path oflength 2n − 2.
The hunter can nowemploy his previousstrategy on this path.This will give O(n log n)capture time.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 133: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/133.jpg)
General graphs
Consider a graph onn vertices.
Pick a spanning tree.
Depth first search yields
This is a closed path oflength 2n − 2.
The hunter can nowemploy his previousstrategy on this path.This will give O(n log n)capture time.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 134: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/134.jpg)
General graphs
Consider a graph onn vertices.
Pick a spanning tree.
Depth first search yields
This is a closed path oflength 2n − 2.
The hunter can nowemploy his previousstrategy on this path.This will give O(n log n)capture time.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 135: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/135.jpg)
General graphs
Consider a graph onn vertices.
Pick a spanning tree.
Depth first search yields
This is a closed path oflength 2n − 2.
The hunter can nowemploy his previousstrategy on this path.This will give O(n log n)capture time.
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 136: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/136.jpg)
On any graph the hunter can catch the rabbit in time O(n log n).
Open Question: If the hunter and rabbit both walk on the samegraph, is the expected capture time O(n)?
Yuval Peres Search Games and Optimal Kakeya Sets
![Page 137: Search Games and Optimal Kakeya Sets Yuval Peres · hhhhhhhhhhhhh hhhhhhhhhhhhhhhhhhhhh Yuval Peres Search Games and Optimal Kakeya Sets. Definition of the game G n Rules At time](https://reader034.vdocument.in/reader034/viewer/2022043010/5f9f61560950b84598734017/html5/thumbnails/137.jpg)
On any graph the hunter can catch the rabbit in time O(n log n).Open Question: If the hunter and rabbit both walk on the samegraph, is the expected capture time O(n)?
Yuval Peres Search Games and Optimal Kakeya Sets