Abstract and subjects
In recent years, the term Cyber Physical Systems (CPS) has gained popularity among researchers and industries. CPS focuses on integrating computation, network, sensor, communication, and control. These devices interact with each other to form a complex system. In CPS, a large number of embedded devices collect and generate data over time for a wide variety of applications. Based on the nature of the application, the devices will have big or real-time data streams. Performing data analysis over such data streams to make future insights or control decisions is an important process to deliver one of the CPS objectives. The machine learning (ML) and deep learning (DL) area have shown cutting edge performance in solving various complex problems. ML and DL algorithms are used as a solution for addressing different security problems in CPS, such as: 1. Anomaly detection, 2. Cybersecurity, 3. Fault prediction, 4. Predictive maintenance, 5. Process optimization, 6. QoS analysis, and 7. Resource allocation. In this chapter, a detailed discussion of ML and DL approaches is presented to address the various challenges posed in CPS such as big data analytics and ways to enable ML/DL capability on the resource constraint devices of CPS.