Predicting Stock Prices with Machine Learning and Virtual Stock Market Game

Predicting Stock Prices with Machine Learning and Virtual Stock Market Game

Priyanshu Mavale, Mohit Sharma, Harshad Phadatare, Prof. Ajitkumar Khachane

Department of Information Technology

Vidyalankar Institute of Technology

Mumbai, India

Abstract- This innovative stock trading simulation platform offers users an immersive experience in the world of stock market investing, using virtual currency. Featuring a user-friendly interface, individuals can easily grasp the essentials of purchasing and selling stocks, managing their investment portfolios, and making informed decisions within a secure, risk-free setting. The virtual currency system ensures that users can refine their trading abilities without any financial jeopardy. Moreover, the simulator delivers real-time market insights, educational materials, and analytical tools to support users in reaching their investment objectives. Our primary project objective is to gamify the experience and foster collaboration among friends within dedicated rooms, enabling them to assess their comparative skills. Users can leverage predictive models, trained using LSTM algorithms, for investment purposes. Extensive testing, employing various prediction algorithms like LSTM, Gradient Boosting Regression, and Random Forest Regression, revealed LSTM's superior performance in stock price forecasting, establishing its efficiency and reliability.

      Keywords—Virtual Currency, Gamification,Prediction Models, Long Short Term Memory (LSTM)