1. ETHEM ALPAYDIN © The MIT Press, 2010 alpaydin@boun.edu.tr http://www.cmpe.boun.edu.tr/~ethem/i2ml2e Lecture Slides for

3. Introduction Modeling dependencies in input; no longer iid Sequences: Temporal: In speech; phonemes in a word (dictionary), words in a sentence (syntax, semantics of the language). In handwriting, pen movements Spatial: In a DNA sequence; base pairs 3Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

4. Discrete Markov Process N states: S 1, S 2,..., S N State at “time” t, q t = S i First-order Markov P(q t+1 =S j | q t =S i, q t-1 =S k,...) = P(q t+1 =S j | q t =S i ) Transition probabilities a ij ≡ P(q t+1 =S j | q t =S i ) a ij ≥ 0 and Σ j=1 N a ij =1 Initial probabilities π i ≡ P(q 1 =S i ) Σ j=1 N π i =1 4Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

5. Stochastic Automaton 5Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

6. Three urns each full of balls of one color S 1 : red, S 2 : blue, S 3 : green Example: Balls and Urns 6Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

7. Given K example sequences of length T Balls and Urns: Learning 7Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

8. Hidden Markov Models States are not observable Discrete observations {v 1,v 2,...,v M } are recorded; a probabilistic function of the state Emission probabilities b j (m) ≡ P(O t =v m | q t =S j ) Example: In each urn, there are balls of different colors, but with different probabilities. For each observation sequence, there are multiple state sequences 8Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

9. HMM Unfolded in Time 9Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

10. Elements of an HMM N: Number of states M: Number of observation symbols A = [a ij ]: N by N state transition probability matrix B = b j (m): N by M observation probability matrix Π = [π i ]: N by 1 initial state probability vector λ = (A, B, Π), parameter set of HMM 10Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

11. Three Basic Problems of HMMs 1. Evaluation: Given λ, and O, calculate P (O | λ) 2. State sequence: Given λ, and O, find Q * such that P (Q * | O, λ ) = max Q P (Q | O, λ ) 3. Learning: Given X={O k } k, find λ * such that P ( X | λ * )=max λ P ( X | λ ) 11 (Rabiner, 1989) Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

12. Forward variable: Evaluation 12Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

13. Backward variable: 13Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

14. Finding the State Sequence 14 No! Choose the state that has the highest probability, for each time step: q t * = arg max i γ t (i) Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

15. Viterbi’s Algorithm δ t (i) ≡ max q1q2∙∙∙ qt-1 p(q 1 q 2 ∙∙∙q t-1,q t =S i,O 1 ∙∙∙O t | λ) Initialization: δ 1 (i) = π i b i (O 1 ), ψ 1 (i) = 0 Recursion: δ t (j) = max i δ t-1 (i)a ij b j (O t ), ψ t (j) = argmax i δ t-1 (i)a ij Termination: p * = max i δ T (i), q T * = argmax i δ T (i) Path backtracking: q t * = ψ t+1 (q t+1 * ), t=T-1, T-2,..., 1 15Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

16. Learning 16Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

17. Baum-Welch (EM) 17 Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

18. Continuous Observations 18 Discrete: Gaussian mixture (Discretize using k-means): Continuous: Use EM to learn parameters, e.g., Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

19. Input-dependent observations: Input-dependent transitions (Meila and Jordan, 1996; Bengio and Frasconi, 1996): Time-delay input: HMM with Input 19Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)

20. Left-to-right HMMs: In classification, for each C i, estimate P (O | λ i ) by a separate HMM and use Bayes’ rule Model Selection in HMM 20Lecture Notes for E Alpaydın 2010 Introduction to Machine Learning 2e © The MIT Press (V1.0)