Minimum classification error training in example based speech and pattern recognition using sparse weight matrices

Authors:
Mike Matton;Dirk Van Compernolle;Ronald Cools
Affiliations:
K.U.Leuven, Department of Computer Science, Celestijnenlaan 200A, B-3001 Heverlee, Belgium;K.U.Leuven, Department ESAT, Kasteelpark Arenberg 10 bus 2441, B-3001 Heverlee, Belgium;K.U.Leuven, Department of Computer Science, Celestijnenlaan 200A, B-3001 Heverlee, Belgium
Venue:
Journal of Computational and Applied Mathematics
Year:
2010

Citing 7
Cited 1

Machine learning, neural and statistical classification

Machine learning, neural and statistical classification
Data Compression and Local Metrics for Nearest Neighbor Classification

IEEE Transactions on Pattern Analysis and Machine Intelligence
A class-dependent weighted dissimilarity measure for nearest neighbor classification problems

Pattern Recognition Letters
An overview of the sparse basic linear algebra subprograms: The new standard from the BLAS technical forum

ACM Transactions on Mathematical Software (TOMS)
Learning Weighted Metrics to Minimize Nearest-Neighbor Classification Error

IEEE Transactions on Pattern Analysis and Machine Intelligence
Learning prototypes and distances: A prototype reduction technique based on nearest neighbor error minimization

Pattern Recognition
Template-Based Continuous Speech Recognition

IEEE Transactions on Audio, Speech, and Language Processing

Time series classification by class-specific Mahalanobis distance measures

Advances in Data Analysis and Classification

Quantified Score

Hi-index	7.30

Visualization

Abstract

The Minimum Classification Error (MCE) criterion is a well-known criterion in pattern classification systems. The aim of MCE training is to minimize the resulting classification error when trying to classify a new data set. Usually, these classification systems use some form of statistical model to describe the data. These systems usually do not work very well when this underlying model is incorrect. Speech recognition systems traditionally use Hidden Markov Models (HMM) with Gaussian (or Gaussian mixture) probability density functions as their basic model. It is well known that these models make some assumptions that are not correct. In example based approaches, these statistical models are absent and are replaced by the pure data. The absence of statistical models has created the need for parameters to model the data space accurately. For this work, we use the MCE criterion to create a system that is able to work together with this example based approach. Moreover, we extend the locally scaled distance measure with sparse, block diagonal weight matrices resulting in a better model for the data space and avoiding the computational load caused by using full matrices. We illustrate the approach with some example experiments on databases from pattern recognition and with speech recognition.