Publication

Abstract : Movement analysis of mentally challenged becomes more open and prevalent with the advancement of sensors embedded in mobile phone devices and cloud infrastructure for data streaming. In this paper, ‘movement analysis of mentally challenged individuals using cloud and machine learning model’ a framework for monitoring specially-abled multiple individuals is proposed. In the proposed framework, different conventional classification techniques of machine learning (ML) algorithms such as logistic classifier, ensemble method, and deep learning (DL) networks such as recurrent neural network (RNN) and dense neural network are used to recognize their activities. The popularly known KAGGLE-UCI datasets which contain smartphone accelerometer data are used to develop the proposed ML model. Performance analysis of classification-based activities recognition schemes is evaluated in terms of precision, recall, F1-score, and accuracy. Results show that DL model with dense and dropout layers after hyperparameter tuning has performed better when compared with rest of the classifiers.