Hybrid Naïve Bayes K-nearest neighbor method implementation on speech emotion recognition

Speech Emotion Recognition technique is incredible in that it can open a way of communication between human and computer. The applications vary from educational software, psychiatric diagnosis, and interrogation to intelligent toys. It has been a long way for researchers who dedicated to search for the best models for speech emotion recognition. This paper proposes a novel hybrid model that combines the K-Nearest Neighbor (KNN) model and the Naïve Bayes (NB) classifier: a model which was inspired from the hybrid model of Support Vector Machine (SVM) and K-Nearest Neighbor method. The implementation of NB-KNN overcomes risks of SVM-KNN model and outperforms the original models that it is composed of.