A study on classification for the vowel signal accordingwith emotional state based on recurent fuzzy C-means algorithm

Abstract

In this paper we used the recurent fuzzy C-means algorithm in order to compare the centers of the clusters with the statistic parameters obtained from the input data sets. Based on the experimental results, the FCM recurent is better than the clasical algorithms FCM and Kmeans. The results are encouraging and justify further research with new features vectors for the automatically emotion recognition.