Elderly Fall Detection Using Machine Learning
Amisha Sinha, Dezy Jha, Malaya Kumar Swain, Soumya Sagar Rath, Nimisha Ghosh · 2024
Elderly fall detection research is aimed at developing algorithms which can be used in different technologies to automatically detect falls and notify caregivers and emergency services. This work uses machine learning algorithms to detect falls. Such detection can greatly reduce the risk of injury in elderly populations. Ongoing research in this area is mainly focused on improving accuracy, reducing false results and developing more accurate detection algorithms. After applying different machine learning algorithms like LDA, CART, Random Forest, Logistic Regression and SVM to calculate average and standard deviation of accuracy, f1-score precision, recall and sensitivity, by running the algorithms 100 times to get the average. Consequently, with an accuracy of 100% the best performance is shown by LDA and SVM.