SVM for automatic rodent sleep-wake classification

Abstract

This work examines the application of a Support Vector Machine (SVM) classifier to automatically detect sleep and wake behaviors in mice based on pressure signals generated by contact with the cage floor. Previous works employed Neural Networks (NN) and Linear Discriminant Analysis (LDA) to classify sleep and wake behaviors. Although the LDA was successful in distinguishing between the sleep and wake behaviors, it has several limitations, which include the need to select a threshold and difficulty separating additional behaviors with subtle differences, such as sleep and rest. The SVM has advantages in that it does not require an independent threshold determination and offers greater degrees of freedom than the LDA for working with complex data sets. In addition, the SVM has direct methods for limiting overfitting for the training sets (unlike the NN method). This paper develops an SVM classifier using a variety of features extracted from the power spectrum, autocorrelation function, and generalized spectrum (autocorrelation of complex spectrum). A genetic algorithm (GA) optimizes the SVM parameters and determines a combination of 5 best features. Experimental results from over 14 hours of data scored by human observation indicate a 95% classification accuracy for SVM.