Self-Organizing Map with Real-time Updating for Big Data Analysis that Uses Bit Value Addition of the RGB Values of the Overlapped Data Points

Abstract

Usually, standard Self-Organizing Maps demand the user to define the number of expected clusters. Most importantly, when there is an update of the data, the data set has to be analyzed using a pre decided algorithm. Thus, it is required to have a high processing capacity to produce real-time analysis of big data. This paper presents a Self-Organizing Maps with Real-time Updating (SOMRU) which eliminates the above-mentioned drawbacks. The proposed SOMRU uses a bitmap as the plotting area.