AI - Based Investment Analysis

This paper presents the completed AI-Based Investment Analyst, a web application developed with React and Vite on the frontend and Python with FastAPI on the backend. The implemented features include a dashboard, stock analysis, risk calculation, forecasting, and FIN-TWIN AI scenario comparison. The frontend and backend were tested together, and the application operated smoothly during the testing carried out. The present results are qualitative; numerical forecasting accuracy and investment-performance results are not claimed because corresponding measurements were not provided. Future work can strengthen the evaluation by documenting the data source, forecasting model, test dataset, evaluation period, and metrics such as MAE and RMSE. Additional improvements may include expanded model comparison, explainability, and more extensive scenario testing.

1.      H. Markowitz, “Portfolio Selection,” The Journal of Finance, vol. 7, no. 1, pp. 77–91, 1952. doi: 10.1111/j.1540-6261.1952.tb01525.x.   2.      T. Fischer and C. Krauss, “Deep learning with long short-term memory networks for financial market predictions,” European Journal of Operational Research, vol. 270, no. 2, pp. 654–669, 2018. doi: 10.1016/j.ejor.2017.11.054.   3.      S. M. Lundberg and S.-I. Lee, “A Unified Approach to Interpreting Model Predictions,” in Advances in Neural Information Processing Systems 30, 2017, pp. 4765–4774.