A Novel Feature Selection Method for Affective Recognition Based on Pulse Signal

Abstract

For the problem of affective recognition from pulse signal, a new feature selection method which combines correlation analysis with max-min ant colony algorithm is proposed in this paper, and stable feature subsets with good performance are found to construct affective recognition model. Firstly, sequential backward selection (SBS) is used for sorting of the original features. Secondly, the linear correlation coefficient is adopted to compute the correlation degrees between the features and features with high correlation degrees are removed through the result of sorting. Finally, max-min ant colony algorithm is used for feature selection, which searches for an optimal subset based on the compact feature subset, and six emotions (happiness, surprise, disgust, grief, anger and fear) are recognized by means of Fisher classifier. The experimental results show that the method can construct effective affective recognition model through stable and effective feature subsets chosen from original features.