Predicting the stock market by using the clustering algorithm in big data

Abstract

As one of the essential approach in record mining and pattern popularity, the Possibilistic C-Means (PCM) algorithm has been widely utilized in evaluation and understanding discovery. It is highly difficult for PCM to provide an awesome end result for clustering huge amount of data particularly for heterogeneous data due to the fact that it is designed for smally established dataset. To address this trouble, we suggest a High-Order PCM (HOPCM). It is a set of rules for massive statistics clustering. The main aim of our proposed system is to find the profit or loss for the clients share based on clustering approach for the specific tickers.