Comparison of Various Feature Selection Algorithms in Speech Emotion Recognition

Speech Emotion Recognition (SER) plays a predominant role in human-machine interaction. SER is challenging task because of number complexities involved it. For an accurate emotion classification system, feature extraction the first and important step carried out on speech signals. And after features are extracted, it very to select best all reject redundant least features. Feature selection methods play performance. The classifier gets selected features, so as reduce unnecessary overload perform better classify emotions. In this study, good combination from Punjabi Emotional Database. Then algorithms explored experimented upon, 1D-CNN used for purpose. results shown compared basis performance metrics. LASSO has other methods.