Body Sensor Cloud Network Based Data Classification By Machine Learning Techniques In Cognitive Human Computer Interaction

Abstract

Abstract Recent developments in cognitive technical systems (CTS), which offer organic and effective operating principles, reveal a development in human-computer interaction (HCI). A CTS must rely on data from several sensors, which must then be processed and merged by fusion algorithms, to do this. To put the observations made into the proper context, additional knowledge sources must also be integrated. This research propose novel technique in cognitive human computer interaction based body sensor data analytics using machine learning technique. here the body sensor based monitoring data has been collected and transmitted by cloud networks for cognitive human computer interaction. then this data has been processed and trained using Boltzmann perceptron basis encoder neural network. Various body sensor-based monitored datasets are subjected to experimental analysis for accuracy, precision, recall, F-1 score, RMSE, normalised square error (NSE), and mean average precision. Proposed technique obtained 93% accuracy, 79% precision, 72% of recall, 64% f-1 score, 51% of RMSE, 56% NSE and 48% MAP.