Construction of Data Driven Decomposition Based Soft Sensors with Auto Encoder Deep Neural Network for IoT Healthcare Applications

The architecture of IoT healthcare is motivated towards the data-driven realization and patient-centric health models, whereas personalized assistance provided by deploying advanced sensors. According to procedures in surgery, emergency unit, patients are monitored till they stable physically then shifted ward for further recovery evaluation. Normally evaluation done doesn’t suggest continuous parameters monitoring physiological condition thus relapse common. In real-time applications, vital will be estimated through dedicated sensors, that still luxurious at present situation highly sensitive harsh conditions environment. Furthermore, monitoring, delay usually Because these issues, soft sensors attractive alternatives. This research this fact Auto Encoder Deep Neural Network (AutoEncDeepNN) proposed depending on Health Framework internet assisting with trigger-based sensor activation model manage master slave advantage method hidden information mined automatically from high representative features generated multiple layer’s iteration. goal consistently achieved outperforms few standard approaches which considered like Hierarchical Extreme Learning Machine (HELM), Convolutional (CNN) Long Short-Term Memory (LSTM). It found AutoEncDeepNN achieves 94.72% accuracy, 41.96% RMSE, 34.16% RAE 48.68% MAE 74.64 ms.