1. Present by: Fang-Hui Chu A Survey of Large Margin Hidden Markov Model Xinwei Li, Hui Jiang York University

2. 2 Reference Papers [Xinwei Li] [M.S. thesis] [Sep. 2005], “Large Margin HMMs for SR” [Xinwei Li] [ICASSP 05], “Large Margin HMMs for SR” [Chaojun Liu] [ICASSP 05], “Discriminative training of CDHMMs for Maximum Relative Separation Margin” [Xinwei Li] [ASRU 05], “A constrained joint optimization method for LME” [Hui Jiang] [SAP 2006], “Large Margin HMMs for SR” [Jinyu Li] [ICSLP 06], “Soft Margin Estimation of HMM parameters”

3. 3 Outline Large Margin HMMs Analysis of Margin in CDHMM Optimization methods for Large Margin HMMs estimation Soft Margin Estimation for HMM

4. 4 Large Margin HMMs for ASR In ASR, given any speech utterance Χ, a speech recognizer will choose the word Ŵ as output based on the plug-in MAP decision rule as follows: For a speech utterance Xi, assuming its true word identity as Wi, the multiclass separation margin for Xi is defined as Discriminant function Ω denotes the set of all possible words

5. 5 Large Margin HMMs for ASR According to the statistical learning theory [Vapnik], the generalization error rate of a classifier in new test sets is theoretically bounded by a quantity related to its margin Motivated by the large margin principle, even for those utterances in the training set which all have positive margin, we may still want to maximize the minimum margin to build an HMM-based large margin classifier for ASR

6. 6 Large Margin HMMs for ASR Given a set of training data D = { X 1, X 2,…,X T }, we usually know the true word identities for all utterances in D, denoted as L = {W 1, W 2,…,W T } First, from all utterances in D, we need to identify a subset of utterances S as We call S as support vector set and each utterance in S is called a support token which has relatively small positive margin among all utterances in the training set D where ε> 0 is a preset positive number

7. 7 Large Margin HMMs for ASR This idea leads to estimating the HMM models Λ based on the criterion of maximizing the minimum margin of all support tokens, which is named as large margin estimation (LME) of HMM The HMM models,, estimated in this way, are called large margin HMMs

8. 8 Analysis of Margin in CDHMM Adopt the Viterbi method to approximate the summation with the single optimal Viterbi path, the discriminant function can be expressed as

9. 9 Analysis of Margin in CDHMM Here, we only consider to estimate mean vectors In this case, the discriminant functions can be represented as a summation of some quadratic terms related to mean values of CDHMMs

10. 10 Analysis of Margin in CDHMM As a result, the decision margin can be represent as a standard diagonal quadratic form Thus, for each feature vector x it, we can divide all of its dimensions into two parts: we can see that each feature dimension contributes to the decision margin separately

11. 11 Analysis of Margin in CDHMM After some math manipulation, we have: linear functionquadratic function

12. 12 Analysis of Margin in CDHMM

13. 13 Analysis of Margin in CDHMM

14. 14 Analysis of Margin in CDHMM

15. 15 Optimization methods for LM HMM estimation An iterative localized optimization method An constrained joint optimization method Semidefinite programming method

16. 16 Iterative localized optimization In order to increase the margin unlimitedly while keeping the margins positive for all samples, both of the models must be moved together –if we keep one of the models fixed, the other model cannot be moved too far under the constraint that all samples must have positive margin –Otherwise the margin for some tokens will become negative Instead of optimizing parameters of all models at the same time, only one selected model will be adjusted in each step of optimization Then the process iterates to update another model until the optimal margin is achieved

17. 17 Iterative localized optimization How to select the target model in each step? –The model should be relevant to the support token with the minimum margin The minimax optimization can be re-formulated as:

18. 18 Iterative localized optimization Approximated by summation of exponential functions

19. 19 Iterative localized optimization

20. 20 Constrained Joint optimization Introduce some constraints to make the optimization problem bounded In this way, the optimization can be performed jointly with respect to all model parameters

21. 21 Constrained Joint optimization In order to bound the margin contribution from the linear part: In order to bound the margin contribution from the quadratic part:

22. 22 Constrained Joint optimization Reformulate the large margin estimation as the following constrained minimax optimization problem:

23. 23 Constrained Joint optimization The constrained minimization problem can be transformed into an unconstrained minimization problem

24. 24 Constrained Joint optimization

25. 25 Soft Margin estimation Model separation measure and frame selection SME objective function and sample selection

26. 26 Soft Margin estimation Difference between SME and LME –LME neglects the misclassified samples. Consequently, LME often needs a very good preliminary estimate from the training set –SME works on all the training data, both the correctly classified and misclassified samples –While SME must first choose a margin ρ heuristically