HMM modification method

• US 20050021337A1
• Filed: 02/24/2004
• Published: 01/27/2005
• Est. Priority Date: 07/23/2003
• Status: Abandoned Application

• Current Assignee: Pantech Co. Ltd. (Pantech Incorporated)
• Original Assignee: Pantech Co. Ltd. (Pantech Incorporated)
• Inventors: Kwon, Tae-Hee
• Application Number:
  US10/787,017
• Publication Number:
  US 20050021337A1
• Time in Patent Office: Days
• US Class Current: 704/256
• CPC Class Codes: G10L 15/144 Training of HMMs

Abstract

A HMM modification method for preventing an overfitting problem, reducing the number of parameters and avoiding gradient calculation by implementing a weighted loss function for misclassification measure and computing a delta coefficient in order to modify a HMM weight is disclosed. The HMM modification method includes the steps of: a) performing Viterbi decoding for pattern classification; b) calculating misclassification measure using discriminant function; c) obtaining modified misclassification measure for a weighted loss function; d) computing a delta coefficient according to the obtained misclassification measure; e) modifying HMM weight according to the delta coefficient; and f) transforming classifier parameters for satisfying a limitation condition.

Claims

1. A HMM modifying method, comprising the steps of:

a) performing Viterbi decoding for pattern classification;

b) calculating misclassification measure using discriminant function;

c) obtaining modified misclassification measure for a weighted loss function;

d) computing a delta coefficient according to the obtained misclassification measure;

e) modifying HMM weight according to the delta coefficient; and

f) transforming classifier parameters for satisfying a limitation condition.

2. The method as recited in claim 1, wherein the weighted loss function {overscore (d)}_i(X;

Λ

) is defined as;

$\begin{array}{c}{\stackrel{\_}{d}}_i<br><br>\left(X;<br><br>\mathrm{\Lambda <br><br>}\right)=<br><br>d_i<br><br>\left(X;<br><br>\mathrm{\Lambda <br><br>}\right)-k·<br><br>g_i<br><br>\left(X;<br><br>\mathrm{\Lambda <br><br>}\right)\\ =<br><br>-\left(1+k\right)·<br><br>g_i<br><br>\left(X;<br><br>\mathrm{\Lambda <br><br>}\right)<br><br>{\log <br><br>\left[\frac{1}{N}<br><br>\underset{j=1,j\ne <br><br>1}{\overset{N}{\sum <br><br>}}<br><br>\exp <br><br>\left[g_j<br><br>\left(X;<br><br>\mathrm{\Lambda <br><br>}\right)<br><br>\mathrm{\eta <br><br>}\right]\right]}^{\frac{1}{\mathrm{\eta <br><br>}}}\end{array}$

, wherein i and j is positive integer number and i representing a number of class, g_i(X;

Λ

) is the discriminant function for class i with Λ

being a set of classifier parameters and X is an observation sequence, N is an integer number representing class models and k is positive number representing the number of HMM state.

3. The method as recited in claim 1, wherein the delta coefficient Δ

w_i is obtained based on the discriminant function and the weighted loss function defined as;

$\mathrm{\Delta <br><br>}<br><br>\text{\hspace{1em}<br><br>}<br><br>w_i=\frac{d<br><br>\text{\hspace{1em}<br><br>}<br><br>i<br><br>\left(X;<br><br>\mathrm{\Lambda <br><br>}\right)}{-g<br><br>\text{\hspace{1em}<br><br>}<br><br>i<br><br>\left(X;<br><br>\mathrm{\Lambda <br><br>}\right)},$

wherein d_i(X;

Λ

) is the weighted loss function and g_i(X;

Λ

) is the discriminant function, Λ

is a set of classifier parameters, X is an observation sequence, i is positive integer number representing a number of class.

4. The method as recited in claim 1, wherein in the step f), the classifier parameter is transformed by the limitation condition, which a summation of HMM weights in a HMM set is limited to a total number of HMM in the HMM set, which is defined as:

$\underset{i=1}{\overset{M}{\sum <br><br>}}<br><br>w_i=M,0<<br><br>w_i<<br><br>M,$

wherein M is positive integer number representing the number of HMM.

5. The method as recited in claim 1, wherein in the step a), the discriminant function is obtained by a viterbi decoding.