Comparative Study of Data Analysis Methods in a Fog

Fog computing is distributed computing consisting of Cloud layer, Fog Layer, and IoT devices/ sensors. Fog computing extends cloud based services closer to end devices in order to provide quick response to real time IoT applications. IoT devices with different sensors generate a huge amount of data which are processed at Fog layer to find insight for business, to reduce workload at Cloud, to reduce use of network bandwidth, to respond in time effective manner or reduce latency and response location awareness. For analysing data effectively at insight for business, Fog layer is a need for intelligent computing systems like data mining and data analytics. In a Fog environment the data analytics can be performed at Fog layer or at Cloud Layer and Fog Layer. The Fog nodes in fog layer are limited in computing resources, so the data analytics must be distributed among the fog nodes to work in a distributed way. Machine learning based data analytics model with various Data analytic methods works better in such situations. In this paper we are comparing various data analytics methods used in Fog computing.