MapReduce with Deep Learning Framework for Student Health Monitoring System using IoT Technology for Big Data

The efficient well-being and health interventions of students are ensured by better knowledge of student's health and fitness factors. Effective Health Monitoring (HM) systems are introduced by using the Internet of Things (IoT) technology and efficient medical services are given by using the personalized health care systems. The sensors used in the IoT may create large amounts of data, which poses computational challenges and data inaccessibility in the IoT environment. Hence, the big data framework has been effectively used in the IoT-based student HM system. Moreover, the characteristics to be accomplished by the big data framework are low-value density, huge data volume, fast update speed, and complex types. This big data framework saves research costs, breaks the traditional space limitations, reflects the true situation of all respondents, and improves research. The shortcomings of conventional student HM systems are effectively solved by this IoT-based big data framework. This paper aims to design an IoT-based student HM system with the utilization of big data and deep learning architecture in order to identify the student's health status. Here, the MapReduce framework is utilized for appropriately processing the big data of student's health. Initially, in the proposed model, the IoT devices are used for collecting the big data through the standard benchmark datasets. The collected big data of student's health are undergone with data normalization technique. In the map phase, the normalized data are given to the Autoencoder and 1-dimension Convolution Neural Network (1DCNN) for extracting the deep features of the student's health information. In the reduce phase, the optimal feature selection is performed with the Adaptive Bird Rat Swarm Optimization (ABRSO) to make the enhancement in the student HM system. Then, the student health status is finally classified with Weight Optimized Recurrent Neural Network (WO-RNN) with Ridge classifier. Here, the optimization takes place in the classification stage using the same ABRSO to achieve superior classification efficiency. The experimental analysis is made to realize the improved performance of the suggested IoT-based student HM system.