Bombay Stock Exchange of India: Patterns and Trends Prediction Using Data Mining Techniques

Introduction

In these days stock market index prediction is an important concern in finance and business but over the years we have been seen it fluctuating over time and most important affects the stock user’s mentality. So determining more effective ways and adopting more and more profitable tools and techniques for stock market index prediction in order to make more accurate decision will be challenging for stock user’s community (Abdoh & Jouhare, 1996). There are several stock exchanges are available in India due to their increasing profit and growing user’s. Bombay Stock Exchange (BSE) is one of the fastest (speed 200 micro seconds) and more prominent stock exchange over the world. It is established in 1875.and has total 5500 companies listed making it world’s no. one exchange in terms of listed companies (Abdulsalam et al., 2011; Ghezelbash, 2012). Over 140 years BSE has become eye witness of growth of Indian corporate sector by providing it an efficient capital raising platform. At the March 2015 end the total market capitalization calculated 1.68 USD Trillion which make it fifth largest leading exchange in the world (Ankerst, 2001). Data mining is a branch of computer science which is designated for finding hidden patterns (previously unknown) from large amount of data (Han & Kamber, 2006). It is one of important step of Knowledge Discovery in Databases (KDD). In today’s competitive world huge amount of information stored in databases. The retrieval and analysis of this information makes data mining important and necessary technique. Financial institution such as stock market produces a vast amount of data throughout the year that provides a road map for data mining tools and techniques and helps to handle complex problems (Fayyad et al., 1996). Over the years various researchers has paid attention of solving these problems and gained a large profit. Obviously this will help the enhancing the data mining research in future. At a time research in data mining is increasing day by day due to the it’s applicability in various fields and increasing information. Data mining in finance specifies various criteria like profit maximization, multi -resolution forecast (days, weekly and yearly) so it will be big challenge to data mining community to predict useful trends and patterns with reasonable cost and short span of time (Fayyad et al., 1996; Han & Kamber, 2006). Since many years stock price prediction is an interesting task for stock users. In the literature number of methods is applied to accomplish this task. There are various approaches of prediction like informal ways (like chart analysis) to formal ways (like regression methods). The only aim of prediction is to reduce uncertainty associated to investment decision making (Boston, 1998).