فهرست مطالب محمد سرچمی

Stock selection and optimal stock portfolio formation depends on various factors which lead to complicate making decision. Investors can maximize investment return and minimize their risk by choosing the optimal stock portfolio. Therefore always they are used to advanced financial algorithms to form the optimal stock portfolio. This study tries to find machine learning and Markowitz models ability in optimal stock portfolio formation and their efficiency comparison. The statistical sample of the present study includes 156 companies listed on the Tehran Stock Exchange during 2010 to 2019. After collecting the data, the deep learning and the Markowitz models were tested using Anaconda software and Python programming language, and then the ability of each model were determined in formation of optimal stock portfolio by portfolio return evaluation criteria, trenors and jensens index.In the ten-share portfolio of the deep learning model; Return of portfolio 0.697, trenors index of 4.541 and Jensens index of 0.480 and in ten-share portfolio of Markowitz model; Portfolio returns of 0.058, trenors index of -1.648 and Jensens index of -0.158 have been calculated. According to the results of the portfolio evaluation, it was concluded that the deep learning model has a higher ability than the Markowitz model in the formation of optimal stock portfolio.

Although econometric models are appropriate for describing and evaluating the relationships between variables and statistical inference, but they have some limitations for financial analysis. Many efforts have been made to model nonlinear relationships in financial data using machine learning technologies. The purpose of this study is to apply machine learning models to form optimal stock portfolios and compare their performance. The statistical sample of the present study consists of 156 companies listed in Tehran Stock Exchange during the period 2009-2018. After data collection, the intended deep learning models in Anaconda software and Python programming language were tested, and then the ability of each model was determined by return evaluation, composite return, trenors and jensens criteria to form an optimal stock portfolio. According to the free-risk and market return rate, forming portfolio by investor to more profit than these two rates and portfolio valuation results of trenors and jensens indexes, it was concluded that the deep Convolutional Neural Network is able to for optimal portfolio. According to this reasoning, the long short-term memory model is not capable of optimal portfolio formation.

Fraud possibility in issued financial statements, negative effects in financial markets and decrease in investment have caused to special attention of responsible regulatory agencies to detection and prevention from fraud. This research aimed to investigation into fuzzy approach ability to fraud detection in financial reporting of some accepted firms by Tehran Stock Exchange. Three hypotheses were stated in this research: 1) Fuzzy decision tree classifier can detect fraud in financial reporting. 2) Sugeno fuzzy classifier can detect fraud in financial reporting. 3) There is a significant difference between results of using fuzzy decision tree classifier and Sugeno fuzzy classifier. The mentioned fuzzy approaches were programmed and hypotheses were tested using Matlab Software. Accuracy average for Fuzzy decision tree classifier was 31/312 and for Sugeno fuzzy classifier was 80/92. It means that the first hypothesis was rejected and the second and third hypotheses were confirmed.