Loading...

Portfolio Recommendation Based on a Hybrid Approach Employing Machine Learning Techniques

Saraee, Arash | 2020

395 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 53397 (01)
  4. University: Sharif University of Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Mostafavi, Mostafa
  7. Abstract:
  8. With the expansion of the use of information technology and public access to financial markets, the number of players in this field has increased and as a result, the nonlinear behavior of the market has become more complex than before. Due to this issue, investors need specific strategies to identify suitable assets for investment and also to determine the time of entry and exit to the stock market in order to make a profitable investment. Therefore, the purpose of this study is to form a stock portfolio as an asset and based on an integrated approach of fundamental and technical analysis using machine learning techniques. In this research, the stock data of Tehran Stock Exchange and OTC Iran in two different time periods of 6 months have been used. Using the CART decision tree model, 6 stock groups were screened based on 12 characteristics of fundamental analysis and 3 criteria of return comparison as a label. Also, based on 12 characteristics of technical analysis and 3 criteria of return comparison as a label, 6 series of trading strategies were developed. Finally, 36 portfolios of all different combinations of characteristics and values of fundamental analysis and trading strategies of technical analysis with specific weights founded and their return results were compared with the criteria. It was shown that the return of the best formed portfolio in the comparison of 3 criteria of return of “The equal weighted index”, “The total index” and “Buy and Hold” strategy of the whole stock group gained more than 222%, 103% and 30% higher returns, respectively
  9. Keywords:
  10. Fundamental Analysis ; Technical Analysis ; Machine Learning ; Stock Portfolio ; Portfolio Recommendation ; Decision Making Tree

 Digital Object List

 Bookmark

No TOC