Human Fall Detection Using One-Class Support Vector Machine Classifier and Feature Ranking
Rashmi Gupta, Jeetendra Kumar · 2025
Identification of unintentional human fall detection is a significant subcategory of human activity recognition. Unintentional falls in older adults cause severe injuries to patients sometimes if a fall is unnoticed, it can cause death also. In this chapter, a feature ranking approach has been proposed to rank the features according to their importance. Accelerometer sensors have been utilized to record activity patterns. For fall detection, three publicly available datasets have been used for training and testing. After data preprocessing, five features have been calculated from the raw data. For classification purposes, one class support vector machine classifier has been used. The proposed feature ranking method can very efficiently rank features according to their importance.