Prescriptive Analytical Models for Dynamic IoT Data Streams:A Review

The application of data analysis tools and procedures to perceive value from vast volume of data created by connected IoT devices is known as IoT data analytics.While predictive analytics on IoT dealing with the prediction involved with the setting of IoT appliances, Prescriptive analytics is the next stage of IoT data analytics involves deriving actionable insights from predictions made in previous stages.The incorporation of time-dependent parameters in prescriptive models provides a more accurate depiction of a complex environment and the decision-making process that goes along with it.The scope of our work is to recommend prescriptive analytical models that make better decisions through the analysis of dynamic IoT data stream in real-time and prescribe an optimal solution.We carry out an analysis of time-series data to identify the patterns of data and learn how they change.In this direction, we attempt to represent time-series data by reducing its length, forecast change points, map change points to prescribed actions, and propose optimal decisions ahead of time events.In this paper, an overview of IoT data analytics, survey of prescriptive analytical models, applications, issues, challenges and platforms for IoT analytics are discussed.