Feature Selection Method for Automatic Sleep Staging based on Genetic Algorithm

Automatic sleep staging monitoring with high accuracy is an effective sleep assessment method and can assist in the treatment of sleep problems. However, multi-domain feature extraction in automatic sleep staging may lead to too many feature parameters, which may cause “dimension curse” in classification. Based on this, genetic algorithm is applied to feature selection of automatic sleep staging. After extracting multiple features from multiple domains of sleep EEG, genetic algorithm is used as a feature selection method to select a feature subset with good resolution, remove redundant features, and improve the efficiency and accuracy of the classifier. The experimental results show that the performance of different classifiers improved after the feature selection of genetic algorithm. As a feature selection method, genetic algorithm can improve the classification accuracy of automatic sleep staging algorithm.