consistencyofstatisticsininfinitedimensionalquotientspaces...

82
Consistency of statistics in infinite dimensional quotient spaces PHD defence of Loïc Devilliers, November, 20, 2017 Prepared at Inria Univeristé Côte d’Azur, CMAP École Polytechnique & ENS Paris-Saclay Jury: Stéphanie Allassonnière Professor Université Paris Descartes Co-advisor Marc Arnaudon Professor Université de Bordeaux Reviewer Charles Bouveyron Professor Université Côte d’Azur President Stephan Huckemann Professor University of Göttingen Reviewer Xavier Pennec Senior Researcher Université Côte d’Azur, Inria Advisor Stefan Sommer Associated Professor University of Copenhagen Reviewer Alain Trouvé Professor ENS Paris-Saclay Examiner 1

Upload: others

Post on 09-Oct-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Consistency of statistics in infinite dimensional quotient spacesPHD defence of Loïc Devilliers, November, 20, 2017

Prepared at Inria Univeristé Côte d’Azur, CMAP École Polytechnique& ENS Paris-Saclay

Jury:Stéphanie Allassonnière Professor Université Paris Descartes Co-advisorMarc Arnaudon Professor Université de Bordeaux ReviewerCharles Bouveyron Professor Université Côte d’Azur PresidentStephan Huckemann Professor University of Göttingen ReviewerXavier Pennec Senior Researcher Université Côte d’Azur, Inria AdvisorStefan Sommer Associated Professor University of Copenhagen ReviewerAlain Trouvé Professor ENS Paris-Saclay Examiner

1

Page 2: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Computational Anatomy: Heart Template Estimation

t0: template, one heart, modeling the others through a diffeomorphism φi.Diffeomorphisms = change the shape but not topology. [Mansi 2009]

(t0, φ1, . . . , φn) = argmint,φ1,...,φn

1n

n∑i=1

(‖t ◦ φi − Patienti‖2 + Regularization(φi)

)2

Page 3: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Computational Anatomy: Brain Template Estimation

[Guimond 1999, Joshi 2004 etc.], Image from [Hamou 2016]

(t0, φ1, . . . , φn) = argmint φ1,...,φn

1n

n∑i=1

(‖t ◦ φi − Yi‖2 + Regularization(φi)

)Template estimation is a tool to statistically analyze diseases.

3

Page 4: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Template Estimation with Surfaces

[courtesy of Pierre Roussillon]

(t0, φ1, . . . , φn) = argmint,φ1,...,φn

1n

n∑i=1

(‖t ◦ φi − Si‖2 + Regularization(φi)

)Goal of this work : study the statistical properties of template estimation. 4

Page 5: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example: Periodic (discretized) signals

Simple example to introduce the Generative Model: In M = Per1(R,R).

0 0.2 0.4 0.6 0.8 1-0.5

0

0.5

1

1.5

Template: t0

5

Page 6: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example: Periodic (discretized) signals

Simple example to introduce the Generative Model: In M = Per1(R,R).

0 0.2 0.4 0.6 0.8 1-0.5

0

0.5

1

1.5

Transformed template by a translation: t0 ◦ ϕ

Note that for the L2 norm, we have ‖t0 ◦ ϕ‖ = ‖t0‖.

5

Page 7: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example: Periodic (discretized) signals

Simple example to introduce the Generative Model: In M = Per1(R,R).

0 0.2 0.4 0.6 0.8 1-0.5

0

0.5

1

1.5

Template and deformed template added to noise: t0 ◦ ϕ+ ε

For instance, Gaussian noise on each point of the discretization grid.Goal: study the statistical properties of the estimator of t0.

5

Page 8: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Generative model

A group G acts on an ambient space M: for g ∈ G, m ∈ M, g · m = gm ∈ M.Observable variable:

Y = Φ · t0 + σε forward modelor

Y = Φ · (t0 + σε) backward model

• Φ a random variable in G.

• t0 the template in M.

• σ > 0 the noise level.

• ε a standardized noise in M: E(ε) = 0, E(‖ε‖2) = 1.

• Φ and ε are independent.

Inverse problem

Given the observed variable Y, how can we estimate the template t0?

6

Page 9: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Generative model

A group G acts on an ambient space M: for g ∈ G, m ∈ M, g · m = gm ∈ M.Observable variable:

Y = Φ · t0 + σε forward modelor

Y = Φ · (t0 + σε) backward model

• Φ a random variable in G.

• t0 the template in M.

• σ > 0 the noise level.

• ε a standardized noise in M: E(ε) = 0, E(‖ε‖2) = 1.

• Φ and ε are independent.

Inverse problem

Given the observed variable Y, how can we estimate the template t0?

6

Page 10: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Minimization (max-max algorithm)

Estimation by minimizing the variance?

The variance at m ∈ M:

F(m) = E(

infg∈G‖m− g · Y‖2 + Regularization(g)

)

The empirical variance at m ∈ M for an n-sample Y1, . . . , Yn:

Fn(m) = infg1,...,gn∈G

(1n

n∑i=1

‖m− gi · Yi‖2

)

Max-max algorithm (also known as Coordinate Descent, GPA, etc.)

Alternatively minimization (over these two steps):

• Step 1: gi← registration of Yi to m, for all i.

• Step 2: m← 1n

n∑i=1

gi · Yi

7

Page 11: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Minimization (max-max algorithm)

Estimation by minimizing the variance?

The variance at m ∈ M:

F(m) = E(

infg∈G‖m− g · Y‖2 + Regularization(g)

)

The empirical variance at m ∈ M for an n-sample Y1, . . . , Yn:

Fn(m) = infg1,...,gn∈G

(1n

n∑i=1

‖m− gi · Yi‖2

)

Max-max algorithm (also known as Coordinate Descent, GPA, etc.)

Alternatively minimization (over these two steps):

• Step 1: gi← registration of Yi to m, for all i.

• Step 2: m← 1n

n∑i=1

gi · Yi

7

Page 12: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example of a failure of max-max algorithm

On the previous example of translated functions: sample of size 105 of discretized functions

with 64 points, σ = 10 [Allassonnière 2007]. Starting point: the template itself.

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2template

Convergence to a local minimum without approximation. 8

Page 13: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example of a failure of max-max algorithm

On the previous example of translated functions: sample of size 105 of discretized functions

with 64 points, σ = 10 [Allassonnière 2007]. Starting point: the template itself.

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2templatecurrent point at the 1th iteration

Convergence to a local minimum without approximation. 8

Page 14: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example of a failure of max-max algorithm

On the previous example of translated functions: sample of size 105 of discretized functions

with 64 points, σ = 10 [Allassonnière 2007]. Starting point: the template itself.

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2templatecurrent point at the 2th iteration

Convergence to a local minimum without approximation. 8

Page 15: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example of a failure of max-max algorithm

On the previous example of translated functions: sample of size 105 of discretized functions

with 64 points, σ = 10 [Allassonnière 2007]. Starting point: the template itself.

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2templatecurrent point at the 3th iteration

Convergence to a local minimum without approximation. 8

Page 16: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example of a failure of max-max algorithm

On the previous example of translated functions: sample of size 105 of discretized functions

with 64 points, σ = 10 [Allassonnière 2007]. Starting point: the template itself.

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2templatecurrent point at the 4th iteration

Convergence to a local minimum without approximation. 8

Page 17: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example of a failure of max-max algorithm

On the previous example of translated functions: sample of size 105 of discretized functions

with 64 points, σ = 10 [Allassonnière 2007]. Starting point: the template itself.

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2templatecurrent point at the 5th iteration

Convergence to a local minimum without approximation. 8

Page 18: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example of a failure of max-max algorithm

On the previous example of translated functions: sample of size 105 of discretized functions

with 64 points, σ = 10 [Allassonnière 2007]. Starting point: the template itself.

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2templatecurrent point at the 10th iteration

Convergence to a local minimum without approximation. 8

Page 19: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example of a failure of max-max algorithm

On the previous example of translated functions: sample of size 105 of discretized functions

with 64 points, σ = 10 [Allassonnière 2007]. Starting point: the template itself.

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2templatecurrent point at the 50th iteration

Convergence to a local minimum without approximation. 8

Page 20: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example of a failure of max-max algorithm

On the previous example of translated functions: sample of size 105 of discretized functions

with 64 points, σ = 10 [Allassonnière 2007]. Starting point: the template itself.

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2templatecurrent point at the 79th iteration

Convergence to a local minimum without approximation. 8

Page 21: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Template estimation with different sample sizes

Starting point: random point

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2

sample size: 2e+05templatemax-max ouput

Inconsistency of the estimator?9

Page 22: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Template estimation with different sample sizes

Starting point: random point

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2

sample size: 4e+05templatemax-max ouput

Inconsistency of the estimator?9

Page 23: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Template estimation with different sample sizes

Starting point: random point

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2

sample size: 6e+05templatemax-max ouput

Inconsistency of the estimator?9

Page 24: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Template estimation with different sample sizes

Starting point: random point

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2

sample size: 8e+05templatemax-max ouput

Inconsistency of the estimator?9

Page 25: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Template estimation with different sample sizes

Starting point: random point

0 0.2 0.4 0.6 0.8 1-1

-0.5

0

0.5

1

1.5

2

sample size: 1e+06templatemax-max ouput

Inconsistency of the estimator?9

Page 26: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Previous works and contributions

Previous works on consistency:

• [Kent & Mardia 1995], [Le 1998] and others restricted to simpletransformations such as rotation, translation, sometimes scaling:

• Consistency with scaling (modification of the algorithm: Y ← Y‖Y‖ ).

• Inconsistency without scaling.

• [Huckemann 2012] Template and estimated template lie ondifferent strata for general action in finite dimensional manifold.

• [Miolane 2017] Consistency Bias = σ2 C2 + o(σ) as σ → 0 in finite

dimensional manifold for Gaussian noise.

Goal of this Phd work: proving and quantifying this inconsistency,in infinite dimensional spaces.

10

Page 27: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Previous works and contributions

Previous works on consistency:

• [Kent & Mardia 1995], [Le 1998] and others restricted to simpletransformations such as rotation, translation, sometimes scaling:

• Consistency with scaling (modification of the algorithm: Y ← Y‖Y‖ ).

• Inconsistency without scaling.

• [Huckemann 2012] Template and estimated template lie ondifferent strata for general action in finite dimensional manifold.

• [Miolane 2017] Consistency Bias = σ2 C2 + o(σ) as σ → 0 in finite

dimensional manifold for Gaussian noise.

Goal of this Phd work: proving and quantifying this inconsistency,in infinite dimensional spaces.

10

Page 28: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Different hypotheses for the action

Isometric Action:

‖g · x‖ = ‖x‖

Invariant Distance:dM(g · x, g · y) = dM(x, y)

General Action

General Action + Regularization Term

What we Want for Application

Part I Part II

The most restrictive hypothesis = the smallest rectangle 11

Page 29: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Table of contents

Introduction

Part I: Inconsistency for Isometric Action

Part II: Inconsistency for Non isometric Action

Conclusion

12

Page 30: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Table of Contents

Introduction

Part I: Inconsistency for Isometric Action

a) Interpretation of the Max-Max Algorithm with the Fréchet Mean inQuotient Spaces

b) Proving the Inconsistency for Isometric Action

c) Quantification of Consistency Bias for Isometric Action

Part II: Inconsistency for Non isometric Action

Conclusion

13

Page 31: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Definitions

Definition of Quotient Space

Orbit of m ∈ M = set of all the points reachable from m:

[m] = {g · m, g ∈ G}.

Quotient space = set of all orbits: Q = M/G = {[m],m ∈ M}.

Definition of Invariant Distance

dM(m,m′) = dM(g · m, g · m′).

Particular case of Invariant Distance: Isometric Action in Hilbert Space

Isometric Action: M a Hilbert, m 7→ g · m linear, ‖g · m‖ = ‖m‖.Proof: ‖g · m− g · m′‖ = ‖g · (m− m′)‖ = ‖m− m′‖.

Classical Proposition: Quotient space = Metric Space

dM invariant quotient distance: dQ ([m], [n]) = infg∈G

dM(m, g · n).

In fact, dQ = pseudo-distance.

14

Page 32: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Definitions

Definition of Quotient Space

Orbit of m ∈ M = set of all the points reachable from m:

[m] = {g · m, g ∈ G}.

Quotient space = set of all orbits: Q = M/G = {[m],m ∈ M}.

Definition of Invariant Distance

dM(m,m′) = dM(g · m, g · m′).

Particular case of Invariant Distance: Isometric Action in Hilbert Space

Isometric Action: M a Hilbert, m 7→ g · m linear, ‖g · m‖ = ‖m‖.Proof: ‖g · m− g · m′‖ = ‖g · (m− m′)‖ = ‖m− m′‖.

Classical Proposition: Quotient space = Metric Space

dM invariant quotient distance: dQ ([m], [n]) = infg∈G

dM(m, g · n).

In fact, dQ = pseudo-distance.

14

Page 33: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Definitions

Definition of Quotient Space

Orbit of m ∈ M = set of all the points reachable from m:

[m] = {g · m, g ∈ G}.

Quotient space = set of all orbits: Q = M/G = {[m],m ∈ M}.

Definition of Invariant Distance

dM(m,m′) = dM(g · m, g · m′).

Particular case of Invariant Distance: Isometric Action in Hilbert Space

Isometric Action: M a Hilbert, m 7→ g · m linear, ‖g · m‖ = ‖m‖.Proof: ‖g · m− g · m′‖ = ‖g · (m− m′)‖ = ‖m− m′‖.

Classical Proposition: Quotient space = Metric Space

dM invariant quotient distance: dQ ([m], [n]) = infg∈G

dM(m, g · n).

In fact, dQ = pseudo-distance.

14

Page 34: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Definitions

Definition of Quotient Space

Orbit of m ∈ M = set of all the points reachable from m:

[m] = {g · m, g ∈ G}.

Quotient space = set of all orbits: Q = M/G = {[m],m ∈ M}.

Definition of Invariant Distance

dM(m,m′) = dM(g · m, g · m′).

Particular case of Invariant Distance: Isometric Action in Hilbert Space

Isometric Action: M a Hilbert, m 7→ g · m linear, ‖g · m‖ = ‖m‖.Proof: ‖g · m− g · m′‖ = ‖g · (m− m′)‖ = ‖m− m′‖.

Classical Proposition: Quotient space = Metric Space

dM invariant quotient distance: dQ ([m], [n]) = infg∈G

dM(m, g · n).

In fact, dQ = pseudo-distance.

14

Page 35: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Fréchet Mean in Metric Spaces

Definition of Fréchet mean in metric spaces

Fréchet Mean of Z a random variable in a metric space (X , dX ):

FM(Z) = argminm∈X

E(d2X (m, Z))

Empirical Fréchet Mean of a n-sample Z1, . . . , Zn:

EFM(Z1, . . . , Zn) = argminm∈X

1n

n∑i=1

d2X (m, Zi)

Example of Hilbert spaces:

For a Hilbert (M, ‖ ‖): FM(Z) = E(Z).

15

Page 36: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Fréchet Mean in Metric Spaces

Definition of Fréchet mean in metric spaces

Fréchet Mean of Z a random variable in a metric space (X , dX ):

FM(Z) = argminm∈X

E(d2X (m, Z))

Empirical Fréchet Mean of a n-sample Z1, . . . , Zn:

EFM(Z1, . . . , Zn) = argminm∈X

1n

n∑i=1

d2X (m, Zi)

Example of Hilbert spaces:

For a Hilbert (M, ‖ ‖): FM(Z) = E(Z).

15

Page 37: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Consistency of Estimation

Fn(m) =1n

n∑i=1

infgi∈G‖m− gi · Yi‖2 =

1n

n∑i=1

d2Q ([m], [Yi])

Minimizing Empirical Variance = Empirical Fréchet Mean (EFM) in Q

Law of large numbers for the sets of (empirical) Fréchet means

Y, (Yn)n i.i.d variables. Thanks to [Ziezold 1977] (if Q is separable):

limn→+∞

EFM([Y1], . . . , [Yn]) ⊂ FM([Y]) a.s.

[t0] not a Fréchet mean of [Y] Inconsistency.

Definition of consistency bias

Consistency bias (CB): distance between [t0] and FM([Y]).

16

Page 38: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Consistency of Estimation

Fn(m) =1n

n∑i=1

infgi∈G‖m− gi · Yi‖2 =

1n

n∑i=1

d2Q ([m], [Yi])

Minimizing Empirical Variance = Empirical Fréchet Mean (EFM) in Q

Law of large numbers for the sets of (empirical) Fréchet means

Y, (Yn)n i.i.d variables. Thanks to [Ziezold 1977] (if Q is separable):

limn→+∞

EFM([Y1], . . . , [Yn]) ⊂ FM([Y]) a.s.

[t0] not a Fréchet mean of [Y] Inconsistency.

Definition of consistency bias

Consistency bias (CB): distance between [t0] and FM([Y]).

16

Page 39: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Consistency of Estimation

Fn(m) =1n

n∑i=1

infgi∈G‖m− gi · Yi‖2 =

1n

n∑i=1

d2Q ([m], [Yi])

Minimizing Empirical Variance = Empirical Fréchet Mean (EFM) in Q

Law of large numbers for the sets of (empirical) Fréchet means

Y, (Yn)n i.i.d variables. Thanks to [Ziezold 1977] (if Q is separable):

limn→+∞

EFM([Y1], . . . , [Yn]) ⊂ FM([Y]) a.s.

[t0] not a Fréchet mean of [Y] Inconsistency.

Definition of consistency bias

Consistency bias (CB): distance between [t0] and FM([Y]).

16

Page 40: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Simple example: the action of rotation

Considering SO(n) acting on Rn by rotation.

•0

••

m

Y dQ ([m], [Y])

Q ' R+

Two orbits (circles), the quotient space (R+), and the distance between orbits

F(m) = E((‖Y‖ − ‖m‖)2), Fréchet mean: ‖m?‖ = E(‖Y‖).Y = Φ · (t0 + σε) ‖m?‖ = E(‖t0 + σε‖)> ‖t0‖ (in general). inconsistency, + Consistency bias computed [Miolane 2017].

Example too simple: infima are removed, not always possible.

17

Page 41: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Simple example: the action of rotation

Considering SO(n) acting on Rn by rotation.

•0

••

m

Y dQ ([m], [Y])

Q ' R+

Two orbits (circles), the quotient space (R+), and the distance between orbits

F(m) = E((‖Y‖ − ‖m‖)2), Fréchet mean: ‖m?‖ = E(‖Y‖).Y = Φ · (t0 + σε) ‖m?‖ = E(‖t0 + σε‖)> ‖t0‖ (in general). inconsistency, + Consistency bias computed [Miolane 2017].

Example too simple: infima are removed, not always possible.

17

Page 42: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Why isometric action is simple?

Our first result of consistency only for isometric action.Isometric action simplification of the square quotient distance:

dQ ([a], [b])2 = infg∈G‖a− g · b‖2 = ‖a‖2 + inf

g∈G(−2 〈a, g · b〉+ ‖g · b‖2)

18

Page 43: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Why isometric action is simple?

Our first result of consistency only for isometric action.Isometric action simplification of the square quotient distance:

dQ ([a], [b])2 = infg∈G‖a− g · b‖2 = ‖a‖2 + inf

g∈G(−2 〈a, g · b〉+ ‖g · b‖2)

= ‖a‖2 + ‖b‖2 + infg∈G

(−2 〈a, g · b〉)

Useful equality for the proof and the quantification of the consistency.

18

Page 44: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Table of Contents

Introduction

Part I: Inconsistency for Isometric Action

a) Interpretation of the Max-Max Algorithm with the Fréchet Mean inQuotient Spaces

b) Proving the Inconsistency for Isometric Action

c) Quantification of Consistency Bias for Isometric Action

Part II: Inconsistency for Non isometric Action

Conclusion

19

Page 45: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Inconsistency for isometric action

0 t0

gt0

g′t0

Cone(t0)

Cone of the template (in gray), and support of t0 + σε (dotted disk).

Theorem: Inconsistency for isometric action in Hilbert space

Observable variable: Y = Φ · (t0 + σε). If:

P(t0 + σε /∈ Cone(t0)) > 0

Then [t0] is not a Fréchet mean of [Y] Inconsistency.20

Page 46: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Sketch of the proof (finite group = more visual proof)

For G finite, R(X) registration of X = t0 + σε to t0.

Gradient of the variance: ∇F(t0) = 2 (E(X)− E(R(X)))

0 t0

gt0

g′t0

Cone(t0)

E(X) = t0

0

•X

•g1X•g2X

•g3X

t0

gt0

g′t0

Cone(t0)

Points in green = Orbit of X.

21

Page 47: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Sketch of the proof (finite group = more visual proof)

For G finite, R(X) registration of X = t0 + σε to t0.

Gradient of the variance: ∇F(t0) = 2 (E(X)− E(R(X)))

0 t0

gt0

g′t0

Cone(t0)

E(X) = t0

0

•X

•R(X)

t0

gt0

g′t0

Cone(t0)

R(X): point in the orbit of X in Cone(t0).

21

Page 48: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Sketch of the proof (finite group = more visual proof)

For G finite, R(X) registration of X = t0 + σε to t0.

Gradient of the variance: ∇F(t0) = 2 (E(X)− E(R(X)))

0 t0

gt0

g′t0

Cone(t0)

E(X) = t0

0

•X

•R(X)

•Xt0

gt0

g′t0

Cone(t0)

X ∈ Cone(t0), then R(X) = X.

21

Page 49: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Sketch of the proof (finite group = more visual proof)

For G finite, R(X) registration of X = t0 + σε to t0.

Gradient of the variance: ∇F(t0) = 2 (E(X)− E(R(X)))

0 t0

gt0

g′t0

Cone(t0)

E(X) = t0

gt0

g′t0

Z0 t0

Cone(t0)

Graphic representation of Z = E(R(X)).The part in grid-line = folded points.

∇F(t0) 6= 0 Inconsistency21

Page 50: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Sketch of the proof (finite or infinite group)

When the group is not finite, differentiate the variance.Two possible methods to show inconsistency:

• Find argmin F, and see if t0 ∈ argmin F : difficult issue.

• Find a point x such has F(x) < F(t0):

We found a point λt0 with F(λt0) < F(t0) Inconsistent.Be careful, a priori [λt0] is not a Fréchet mean of [Y].

22

Page 51: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

How often is fulfilled this condition with the cone?

A group G acts isometrically on a Hilbert space. [t0] a manifold,Tt0 [t0] the affine tangent space of [t0] at t0.Tt0 [t0]⊥ the normal space of [t0] at t0.

Proposition: being inconsistent for smooth orbits.

P(ε /∈ Tt0 [t0]⊥) > 0 =⇒ inconsistency

[t0]

Tt0 [t0]⊥

Tt0 [t0]

g · t0

0t0

y

y /∈ Tt0 [t0]⊥ therefore y is closer from g · t0 for some g ∈ G than t0 itself. In

conclusion, y in the support of X = t0 + σε inconsistency.

23

Page 52: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Table of Contents

Introduction

Part I: Inconsistency for Isometric Action

a) Interpretation of the Max-Max Algorithm with the Fréchet Mean inQuotient Spaces

b) Proving the Inconsistency for Isometric Action

c) Quantification of Consistency Bias for Isometric Action

Part II: Inconsistency for Non isometric Action

Conclusion

24

Page 53: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Consistency bias when the noise level tends to infinity

Definition of consistency bias

Consistency bias (CB) : distance between the template t0 and argmin F.

Definition of fixed points

A fixed point m ∈ M : for all g ∈ G, g · m = m.

Proposition: consistency bias is asymptotically linear when σ → +∞G acts isometrically on a Hilbert space. We take Y = Φ · t0 + σε.If support of the noise ε is not included in the set of fixed points then:

CB = σK + o(σ) as σ → +∞, where K = sup‖v‖=1

E

(supg∈G〈v, g · ε〉

)> 0.

Moreover, limt0→0

CB = σK.

25

Page 54: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Consistency bias when the noise level tends to infinity

Definition of consistency bias

Consistency bias (CB) : distance between the template t0 and argmin F.

Definition of fixed points

A fixed point m ∈ M : for all g ∈ G, g · m = m.

Proposition: consistency bias is asymptotically linear when σ → +∞G acts isometrically on a Hilbert space. We take Y = Φ · t0 + σε.If support of the noise ε is not included in the set of fixed points then:

CB = σK + o(σ) as σ → +∞, where K = sup‖v‖=1

E

(supg∈G〈v, g · ε〉

)> 0.

Moreover, limt0→0

CB = σK.

25

Page 55: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Sketch of the proof

F(m) = E(

infg∈G‖m− g · Y‖2

)where Y =Φ · t0 +σε.

• Minimization of F(λv) w.r.t. λ ≥ 0, ‖v‖ = 1. Then m? ∈ argmin F

‖m?‖ = sup‖v‖=1

E

(supg∈G〈v, g · Y〉

)

= sup‖v‖=1

E

(supg∈G

(〈v, gΦt0〉+ 〈v, σgε〉)

)Difficult (impossible?) to compute.

• Cauchy-Schwarz inequality:

−‖t0‖+ σK ≤ ‖m?‖ ≤ ‖t0‖+ σK

• By triangular inequality:

−2‖t0‖+ σK ≤ ‖m? − t0‖ ≤ σK + 2‖t0‖

K > 0 (because the support ε is not included in the set of fixed points).

26

Page 56: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Sketch of the proof

F(m) = E(

infg∈G‖m− g · Y‖2

)where Y =Φ · t0 +σε.

• Minimization of F(λv) w.r.t. λ ≥ 0, ‖v‖ = 1. Then m? ∈ argmin F

‖m?‖ = sup‖v‖=1

E

(supg∈G〈v, g · Y〉

)= sup‖v‖=1

E

(supg∈G

(〈v, gΦt0〉+ 〈v, σgε〉)

)Difficult (impossible?) to compute.

• Cauchy-Schwarz inequality:

−‖t0‖+ σK ≤ ‖m?‖ ≤ ‖t0‖+ σK

• By triangular inequality:

−2‖t0‖+ σK ≤ ‖m? − t0‖ ≤ σK + 2‖t0‖

K > 0 (because the support ε is not included in the set of fixed points).

26

Page 57: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Sketch of the proof

F(m) = E(

infg∈G‖m− g · Y‖2

)where Y =Φ · t0 +σε.

• Minimization of F(λv) w.r.t. λ ≥ 0, ‖v‖ = 1. Then m? ∈ argmin F

‖m?‖ = sup‖v‖=1

E

(supg∈G〈v, g · Y〉

)= sup‖v‖=1

E

(supg∈G

(〈v, gΦt0〉+ 〈v, σgε〉)

)Difficult (impossible?) to compute.

• Cauchy-Schwarz inequality:

−‖t0‖+ σK ≤ ‖m?‖ ≤ ‖t0‖+ σK

• By triangular inequality:

−2‖t0‖+ σK ≤ ‖m? − t0‖ ≤ σK + 2‖t0‖

K > 0 (because the support ε is not included in the set of fixed points).

26

Page 58: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Sketch of the proof

F(m) = E(

infg∈G‖m− g · Y‖2

)where Y =Φ · t0 +σε.

• Minimization of F(λv) w.r.t. λ ≥ 0, ‖v‖ = 1. Then m? ∈ argmin F

‖m?‖ = sup‖v‖=1

E

(supg∈G〈v, g · Y〉

)= sup‖v‖=1

E

(supg∈G

(〈v, gΦt0〉+ 〈v, σgε〉)

)Difficult (impossible?) to compute.

• Cauchy-Schwarz inequality:

−‖t0‖+ σK ≤ ‖m?‖ ≤ ‖t0‖+ σK

• By triangular inequality:

−2‖t0‖+ σK ≤ ‖m? − t0‖ ≤ σK + 2‖t0‖

K > 0 (because the support ε is not included in the set of fixed points).

26

Page 59: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Table of Contents

Introduction

Part I: Inconsistency for Isometric Action

Part II: Inconsistency for Non isometric Action

a) Inconsistency for Invariant Distance

b) Inconsistency for Non Invariant Distance

Conclusion

27

Page 60: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Variation of the Isotropy Group Due to the Noise

Definition: Isotropy Group (or Stabilizer)

Iso(m) = {g ∈ G, s.t. g · m = m}

Example: Reparametrization of functions

ϕ : [0, 1]→ [0, 1] homeomorphism, f : [0, 1]→ R (ϕ, f) 7→ ϕ · f = f ◦ ϕ

t0 constant map on D = [0.2, 0.8]

0 0.2 0.4 0.6 0.8 10

0.1

0.2

0.3

0.4

0.5t0

Iso(t0) = {ϕ | ϕ|Dc = Id} ! {Id}

t0 + noise

0 0.2 0.4 0.6 0.8 10

0.1

0.2

0.3

0.4

0.5t0+noise

Iso(t0 + noise) = {Id}

28

Page 61: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Variation of the Isotropy Group Due to the Noise

Definition: Isotropy Group (or Stabilizer)

Iso(m) = {g ∈ G, s.t. g · m = m}

Example: Reparametrization of functions

ϕ : [0, 1]→ [0, 1] homeomorphism, f : [0, 1]→ R (ϕ, f) 7→ ϕ · f = f ◦ ϕ

t0 constant map on D = [0.2, 0.8]

0 0.2 0.4 0.6 0.8 10

0.1

0.2

0.3

0.4

0.5t0

Iso(t0) = {ϕ | ϕ|Dc = Id} ! {Id}

t0 + noise

0 0.2 0.4 0.6 0.8 10

0.1

0.2

0.3

0.4

0.5t0+noise

Iso(t0 + noise) = {Id}28

Page 62: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Stability Theorem Implies Inconsistency

Stability Theorem in Hilbert spaces

G a compact group acting continuously on M a Hilbert space, dM isinvariant. Observable variable Y in M. If

P(Iso(Y) = {eG}) > 0 eG : neutral element in G.

m? ∈ argminm∈M

F(m) = argminm∈M

E(

infg∈G

dM(m, g · Y)2

).

If R(Y) is a measurable variable registering Y to m?, then:

Iso(m?) = {eG}.

Implies Inconsistency if Iso(t0) 6= {eG}.Stability Theorem also true in complete finite dimensional Riemannianmanifolds and proof of the measurable variable R(Y) [Huckemann 2012].

29

Page 63: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Table of Contents

Introduction

Part I: Inconsistency for Isometric Action

Part II: Inconsistency for Non isometric Action

a) Inconsistency for Invariant Distance

b) Inconsistency for Non Invariant Distance

Conclusion

30

Page 64: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Non Invariant Distance

Non invariant distance used in applications:

Reparametrization by a diffeomorphism ϕ

fi : Rd → R images d = 2 or signals d = 1: ‖f1 ◦ϕ− f2 ◦ϕ‖2 6= ‖f1− f2‖2.

G acting on a Hilbert space:A priori, possibility to define a distance in the quotient space.For Y = Φ · t0 + σε. minimizing F(m) = E( inf

g∈G‖Y − g · m‖2): still possible.

31

Page 65: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Non Invariant Distance

Non invariant distance used in applications:

Reparametrization by a diffeomorphism ϕ

fi : Rd → R images d = 2 or signals d = 1: ‖f1 ◦ϕ− f2 ◦ϕ‖2 6= ‖f1− f2‖2.

G acting on a Hilbert space:A priori, possibility to define a distance in the quotient space.For Y = Φ · t0 + σε. minimizing F(m) = E( inf

g∈G‖Y − g · m‖2): still possible.

31

Page 66: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

How to deal with non isometric action?

Isometric

σ•t0•0

Orbit of the template, in gray the noise.

We can find a point λt0 such thatF(λt0) < F(t0).

General Action with Bounded Orbit

•t0•0

σ

Bounded orbit of the template, in graythe noise.

32

Page 67: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

How to deal with non isometric action?

Isometric

σ•t0•0

Orbit of the template, in gray the noise.

We can find a point λt0 such thatF(λt0) < F(t0).

General Action with Bounded Orbit

•t0•0

σ

Bounded orbit of the template, in graythe noise.

32

Page 68: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

How to deal with non isometric action?

Isometric

σ•0

Orbit of the template, in gray the noise.

We can find a point λt0 such thatF(λt0) < F(t0).

General Action with Bounded Orbit

σ•0

Bounded orbit of the template, in graythe noise.

So why not in this case?

32

Page 69: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Inconsistency for non invariant distance

Inconsistency: a subgroup of G acts isometrically

A group G acting on a Hilbert space, [t0] is bounded. We note:

θ(G) =1‖t0‖

E

(supg∈G〈g · t0, ε〉

)

If H a subgroup of G, H acts isometrically and θ(H) > 0,then inconsistency for σ > σc = f([t0], θ(G), θ(H), t0) for a certainpositive function f .

Example

G = group of diffeomorphisms, H = rotations.

33

Page 70: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Inconsistency for non invariant distance

Inconsistency: a subgroup of G acts isometrically

A group G acting on a Hilbert space, [t0] is bounded. We note:

θ(G) =1‖t0‖

E

(supg∈G〈g · t0, ε〉

)

If H a subgroup of G, H acts isometrically and θ(H) > 0,then inconsistency for σ > σc = f([t0], θ(G), θ(H), t0) for a certainpositive function f .

Example

G = group of diffeomorphisms, H = rotations.

33

Page 71: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Inconsistency for non invariant distance

Inconsistency for G acting linearly + Regularization

A group G acting linearly on a Hilbert space, [t0] is bounded. We note:

θ(G) =1‖t0‖

E

(supg∈G〈g · t0, ε〉

).

The template estimation is performed by minimizing

F(m) = E(

infg∈G‖g · m− Y‖2 + Regularization(g)

),

where Regularization is bounded. If θ(G) > 0 then Inconsistency forσ > σc = f([t0], θ(G), t0) for a certain positive function f .

Action of reparametrization of functions

ϕ a diffeo (ϕ, f) 7→ f ◦ ϕ linear action.Proof: (af1 + f2) ◦ ϕ = af1 ◦ ϕ + f2 ◦ ϕ.

34

Page 72: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Inconsistency for non invariant distance

Inconsistency for G acting linearly + Regularization

A group G acting linearly on a Hilbert space, [t0] is bounded. We note:

θ(G) =1‖t0‖

E

(supg∈G〈g · t0, ε〉

).

The template estimation is performed by minimizing

F(m) = E(

infg∈G‖g · m− Y‖2 + Regularization(g)

),

where Regularization is bounded. If θ(G) > 0 then Inconsistency forσ > σc = f([t0], θ(G), t0) for a certain positive function f .

Action of reparametrization of functions

ϕ a diffeo (ϕ, f) 7→ f ◦ ϕ linear action.Proof: (af1 + f2) ◦ ϕ = af1 ◦ ϕ + f2 ◦ ϕ.

34

Page 73: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Table of contents

Introduction

Part I: Inconsistency for Isometric Action

Part II: Inconsistency for Non isometric Action

Conclusion

35

Page 74: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Summary of contributions

• It is proved that the template estimation with the Fréchet mean inquotient space is not consistent for isometric action.

• It is possible to quantify the consistency bias for σ → +∞.

• We proved a stability theorem which implies the inconsistency inHilbert Space for invariant distance.

• The inconsistency can also be proved for not isometric action, but onlyfor σ high enough.

This work has been presented in a workshsop (MFCA 2015), published in aconference (IPMI 2017) and in two journal papers (SIIMS 2017 and Entropy2017).

36

Page 75: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Summary of contributions

• It is proved that the template estimation with the Fréchet mean inquotient space is not consistent for isometric action.

• It is possible to quantify the consistency bias for σ → +∞.

• We proved a stability theorem which implies the inconsistency inHilbert Space for invariant distance.

• The inconsistency can also be proved for not isometric action, but onlyfor σ high enough.

This work has been presented in a workshsop (MFCA 2015), published in aconference (IPMI 2017) and in two journal papers (SIIMS 2017 and Entropy2017).

36

Page 76: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

What are the possible extensions?

• Extending the existence of the measurable variable which registers datato a certain point.

• Proving the inconsistency for non invariant distance for all σ.

• Provide an asymptotic behaviour of the consistency bias when σ → 0.

37

Page 77: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Thank you for your attention!Any questions?

37

Page 78: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example 2: action of diffeomorphisms on functions

0 0.2 0.4 0.6 0.8 1-2

-1

0

1

2

Template: t0

Page 79: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example 2: action of diffeomorphisms on functions

0 0.2 0.4 0.6 0.8 1-2

-1

0

1

2

Deformed template: t0 ◦ ϕ

SRVF: The norm of f√|f|

is invariant under the action of ϕ commonly used.

Page 80: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example 2: action of diffeomorphisms on functions

0 0.2 0.4 0.6 0.8 1-2

-1

0

1

2

Template and deformed template added to noise: t0 ◦ ϕ+ ε

Page 81: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example 3: Consistency and smoothness

Example of translated functions: sample size 106 of discretized functionswith 64 points, σ = 10.

0 0.2 0.4 0.6 0.8 1-0.5

0

0.5

1

1.5templatemax max output

Page 82: Consistencyofstatisticsininfinitedimensionalquotientspaces …loic.devilliers.free.fr/beamerLD.pdf · 3. TemplateEstimationwithSurfaces [courtesy of Pierre Roussillon] (bt 0;˚^

Example 4: Local minima

0.4 0.45 0.5 0.55 0.6

1

1.2

1.4