PERFORMANCE EVALUATION OF DIFFERENT WEIGHTING SCHEMES ON KNN-BASED EMOTION RECOGNITION IN MANDARIN SPEECH

Abstract

Since emotion is important in influencing cognition, perception of daily activities such as learning, communication and even rational decision-making, it must be considered in human-computer interaction. In this paper, we compare four different weighting functions in weighted KNN-based classifiers to recognize five emotions, including anger, happiness, sadness, neutral and boredom, from Mandarin emotional speech. The classifiers studied include weighted KNN, weighted CAP, and weighted D-KNN. We use the result of traditional KNN classifier as the line performance measure. The experimental results show that the used Fibonacci weighting function outperforms others in all weighted classifiers. The highest accuracy achieves 81.4% with weighted D-KNN classifier.