Association Rule Mining for Stock Data

Abstract

Stock market is a place where erratic change occurs, therefore there is a need to discover some most appropriate rules or associations which would help the investors/traders to make accurate decisions for the investment in the stock market.This study uses Apriori Algorithm forAssociation Rule Mining to predict best association rules on the stock dataset from National Stock Exchange, India. Theoretical and experimental methodology is adopted to achieve the objectives of the study. Algorithm is implemented on six months pre-processed stock dataset using WEKA tool. Frequent itemset with size 3 and six best association rules are predicted. The study shows that maximum association rules generated are identical to the patterns of the frequent itemset.

Keywords:Association Rule Mining, Stock Market, Weka, Frequent Itemset, Support, Confidence, National Stock Exchange, Data Mining, Apriori Algorithm