Using Adaptive Genetic Algorithms to Improve Speech Emotion Recognition

In this paper, adaptive genetic algorithms are employed to search for the worst performing features with respect to the probability of correct classification achieved by the Bayes classifier in a first stage. These features are subsequently excluded from sequential floating feature selection that employs the probability of correct classification of the Bayes classifier as criterion. In a second stage, adaptive genetic algorithms search for the worst performing utterances with respect to the same criterion. The sequential application of both stages is demonstrated to improve speech emotion recognition on the Danish Emotional Speech database.