Big Data Stream Discretization Using Chimerge Algorithm

In this paper, we propose a new approach of the ChiMerge algorithm. We have added a new layer to remediate the threshold limitation issue and process the incoming data in real-time with a minimal latency. Our empirical results show that ChiMerge stream discretization helps to improve the subsequent pre-processing and algorithm execution time by 3.5x compared to batch processing, with a slightly different accuracy rate.