Importance of the bioradar signal preprocessing in fall detection

This paper describes results of fall detection study with a non-contact bioradiolo-cation method. The exploration of the impact of the accuracy of determining event length on classification is discussed. In order to investigate this question an analysis including the machine learning classification was performed. We compare performance of two machine learning algorithms: the support vector machines and “K-nearest neighbors” method. The results demonstrated that inaccuracy in determination of length of event has a minor influence on the final classification performance.