Providing an intelligent recommender system to increase the profitability of investors in the Tehran Stock Exchange

Main Article Content

Edris Nanvai Savojbolaghy, et. al.


Recommender systems are the most important tools of data mining, which have been today created to predict the profitability in the stock exchange and providing relevant information. These systems can collect the data related to the stock exchange and the status per share from different references explicitly or implicitly and use them to predict and increase the earnings of investors. The investors used to apply return on investment (ROI) ratios to make financial decisions. Although the ratios gained pervasive function in practice, they never considered the time value of money (TVM) and investment risk. Later, the concept of TVM entered into the literature of the financial and investment economy and left significant effects on financial decisions. Accordingly, this study has proposed a recommender system to enhance the profitability of investors in the Tehran Stock Exchange. To do so, the combination of the NSGAII algorithm and clustering technique based on the K-Means algorithm was used. NSGAII algorithm can be used in this solution to derive the main specifications from the dataset to improve the efficiency and accuracy of clustering. In this case, the specifications with no impact on the prediction of profitability are eliminated, and the clustering is done on the main specifications without outliers. The process of elimination of outliers would be completed without losing the basic data of the dataset. Therefore, a purified set of rules is obtained by narrowing down the data, which can facilitate decision-making. In the end, the proposed solution was implemented in MATLAB based on Crisp Methodology. Using the solution, the risky, low-risk, and moderate-risk portfolios can be extracted. The purchase volume from the symbols of the five companies listed in the stock exchange differs based on the risk level and expected return.

Article Details