The past few decades have witnessed a sharp increase in life expectancy. As a result, the proportion of elderly people is increasing worldwide. Consequently, Dementia and Parkinson's disease are expected to rise, thereby increasing the risk of critical events such as falls for elderly people. This has prompted many researchers to develop a wide range of solutions for fall detection and prevention. However, these solutions are either inaccurate or impractical due to hardware complexity. In this paper, we have proposed a novel Random Neural Network (RNN) based fall detection scheme. Results obtained from the proposed RNN-based scheme are compared with traditional machine learning methods such as Support Vector Machine (SVM) and traditional Artificial Neural Network (ANN) etc. From the results, it is evident that the proposed scheme has a higher accuracy of 98%. Additionally, several other parameters such as precision, recall, specificity, and F-measure show that the proposed algorithm has better generalisation capabilities when compared with other traditional machine learning schemes. Furthermore, the proposed RNN is also compared with a recent scheme and the obtained results demonstrate the superiority of the proposed scheme.
- fall detection, remote healthcare, neural networks, machine learning, RNN