A comparative study of SVM kernel applied to emotion recognition from physiological signals

This paper investigates the performance of support vector machines with linear, cubic and radial basis function (RBF) kernels in the problem of emotion recognition from physiological signals. Five physiological signals: blood volume pulse (BVP), electromyography (EMG), skin conductance (SC), skin temperature (SKT) and respiration (RESP) were selected to extract 30 features for recognition. Support vector machine(SVM) is a new technique for pattern classification, and is used in many applications. Kernel type in the SVM training process, along with feature selection, will significantly impact classification accuracy. Experiments are designed and carried out to find the best SVM kernel among linear, cubic, and RBF for emotions recognition. The experimental results indicate that the proposed method provides very stable and successful emotional classification performance over six emotional states.