IoT and Deep Learning on Sensor Data
Abdul Quadir, Siva AN, Arun Kumar Sivaraman · Zenodo (CERN European Organization for Nuclear Research) · 2021
Internet of things (IoT) combined with machine learning algorithms has several applications in various fields. Extensive research and projects are available in Intrusion detection systems (IDS). This paper focuses on collecting data from various sensors and storing them efficiently in various database systems and applying machine learning algorithms to extract useful information from raw sensor data. We propose a 3-tier system that consists of sensor nodes, a sensor database, combined with machine learning algorithm for predicting the activity performed by the human at that point in time. The dataset used to train the Machine learning algorithm has been collected and stored in a NoSQL database which is non-relational and easily scalable. This work compares the reliability and accuracy of the existing algorithms in the field of IoT in sensor data.