The application of data mining techniques in analysis the stock portfolio in order to identify common patterns in the behavior of shareholders (Case study of selected brokers in Mazandaran province)

In this study, we examined the analysis of stock portfolio stakeholders in order to identify common patterns in the behavior of shareholders. Information required about the shareholders shopping cart was collected from the selected brokers in Mazandaran province / Sari city. This information includes demographic information, such as gender, occupation and education, along with a basket of shares purchased during 2013. Data were collected for 150 shares that during this period have traded at least 5 times. This study for the first time with the use of information about the stock portfolio shareholders has offered a model to identify the pattern of purchase behavior of shareholders. In the proposed model, at first shareholders, according to the number of shares purchased categorized into three levels of risk appetite (high, medium, low) then, with the help of data mining classification techniques and association rules have been extracted rules that can be used as a model to predict the behavior of shareholders.