hidden markov models tunghai university fall 2005

16
Hidden Markov Models Tunghai University Fall 2005

Upload: emily-price

Post on 18-Dec-2015

216 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Hidden Markov Models Tunghai University Fall 2005

Hidden Markov Models

Tunghai University

Fall 2005

Page 2: Hidden Markov Models Tunghai University Fall 2005

Simple Model - Markov Chains

• Markov Property: The state of the system at time t+1 only depends on the state of the system at time t

X1X2 X3 X4 X5

] x X | x P[X

] x X , x X , . . . , x X , x X | x P[X

tt11t

00111-t1-ttt11t

t

t

Page 3: Hidden Markov Models Tunghai University Fall 2005

Markov Chains

Stationarity Assumption

• Probabilities are independent of t when the process is

“stationary”

So,

This means that if system is in state i, the probability that

the system will transition to state j is pij no matter what

the value of t is

pij ] x X| x P[X itj1t

Page 4: Hidden Markov Models Tunghai University Fall 2005

Weather:

– raining today rain tomorrow prr = 0.4

– raining today no rain tomorrow prn = 0.6

– no raining today rain tomorrow pnr = 0.2

– no raining today no rain tomorrow prr = 0.8

Simple Example

Page 5: Hidden Markov Models Tunghai University Fall 2005

Simple Example

Transition Matrix for Example

• Note that rows sum to 1

• Such a matrix is called a Stochastic Matrix

• If the rows of a matrix and the columns of a matrix all sum to 1, we have a Doubly Stochastic Matrix

8.02.0

6.04.0P

Page 6: Hidden Markov Models Tunghai University Fall 2005

Gambler’s Example

– At each play we have the following:

• Gambler wins $1 with probability p

• Gambler loses $1 with probability 1-p

– Game ends when gambler goes broke, or gains a fortune of $100

– Both $0 and $100 are absorbing states

0 1 2 N-1 N

p p p p

1-p 1-p 1-p 1-pStart (10$)

or

Page 7: Hidden Markov Models Tunghai University Fall 2005

Coke vs. Pepsi

Given that a person’s last cola purchase was Coke, there is a 90% chance that her next cola purchase will also be Coke.

If a person’s last cola purchase was Pepsi, there is an 80% chance that her next cola purchase will also be Pepsi.

coke pepsi

0.10.9 0.8

0.2

Page 8: Hidden Markov Models Tunghai University Fall 2005

Coke vs. Pepsi

Given that a person is currently a Pepsi purchaser, what is the probability that she will purchase Coke two purchases from now?

66.034.0

17.083.0

8.02.0

1.09.0

8.02.0

1.09.02P

8.02.0

1.09.0P

The transition matrix is:

(Corresponding to one purchase ahead)

Page 9: Hidden Markov Models Tunghai University Fall 2005

Coke vs. Pepsi

Given that a person is currently a Coke drinker, what is the probability that she will purchase Pepsi three purchases from now?

562.0438.0

219.0781.0

66.034.0

17.083.0

8.02.0

1.09.03P

Page 10: Hidden Markov Models Tunghai University Fall 2005

Coke vs. Pepsi

Assume each person makes one cola purchase per week. Suppose 60% of all people now drink Coke, and 40% drink Pepsi.

What fraction of people will be drinking Coke three weeks from now?

6438.0438.04.0781.06.0)0( )3(101

)3(000

1

0

)3(03

pQpQpQXPi

ii

Let (Q0,Q1)=(0.6,0.4) be the initial probabilities.

We will regard Coke as 0 and Pepsi as 1

We want to find P(X3=0)

8.02.0

1.09.0P

P00

Page 11: Hidden Markov Models Tunghai University Fall 2005

Hidden Markov Models - HMM

H1 H2 HL-1 HL

X1 X2 XL-1 XL

Hi

Xi

Hidden variables

Observed data

Page 12: Hidden Markov Models Tunghai University Fall 2005

Coin-Tossing Example

0.9

Fair loaded

head head

tailtail

0.9

0.1

0.1

1/2 1/4

3/41/2

H1 H2 HL-1 HL

X1 X2 XL-1 XL

Hi

Xi

L tosses Fair/Loaded

Head/Tail

Start

1/2 1/2

Page 13: Hidden Markov Models Tunghai University Fall 2005

H1 H2 HL-1 HL

X1 X2 XL-1 XL

Hi

Xi

L tosses

Fair/Loaded

Head/Tail

0.9

Fair loaded

head head

tailtail

0.9

0.1

0.1

1/2 1/4

3/41/2

Start1/2 1/2

Coin-Tossing Example

Query: what are the most likely values in the H-nodes to generate the given data?

Page 14: Hidden Markov Models Tunghai University Fall 2005

1. Compute the posteriori belief in Hi (specific i) given the evidence {x1,…,xL} for each of Hi’s values hi, namely, compute p(hi | x1,…,xL).

2. Do the same computation for every Hi but without repeating the first task L times.

Coin-Tossing Example

Seeing the set of outcomes {x1,…,xL}, compute p(loaded | x1,…,xL) for each coin toss

Query: what are the probabilities for fair/loaded coins given the set of outcomes {x1,…,xL}?

Page 15: Hidden Markov Models Tunghai University Fall 2005

C-G Islands Example

Regular

DNA

C-G island

C-G islands: DNA parts which are very rich in C and G

A

C

G

T

change

A

C

G

T

(1-P)/4

P/6

q/4

q/4

q/4

q/4 P

P

q

q

qqP

P

(1-q)/6

(1-q)/3

p/3

p/3

p/6

Page 16: Hidden Markov Models Tunghai University Fall 2005

C-G Islands Example

A

C

G

T

change

A

C

G

T

H1 H2 HL-1 HL

X1 X2 XL-1 XL

Hi

Xi

C-G island?

A/C/G/T