Supervisor Faculty Member

Shervin Rahimzadeh Arashloo

Github Link

About Project

Stock Vision is a mobile application designed to provide users with a real-time stock/ETF simulation using past stock/ETF data. It allows users to invest using our application’s fake currency and get real-time results/tips about their investments. The application leverages advanced machine learning algorithms and data analysis techniques to forecast the future trends of various stocks and teaches users the fundamentals of stock/ETF predictions with hands-on experience simulation.

The application employs cutting-edge algorithms to capture complex patterns and correlations within stock data. Users capture different patterns for different time range graphs, such as daily, weekly, monthly, or annually. There are 9 fundamental patterns captured by our algorithms, such as doubles, flag, head and shoulders, rectangle, rounding bottom, support and resistance, triangle, triples, and wedge. After analyzing graphs and capturing complex patterns, a unique machine-learning technique, ARIMA, is used to predict results and provide accurate and up-to-date tips. After implementing LSTM, RNN, and Logistic Regression, ARIMA gave the most accurate results and was used in the application. The accuracy of the predictions is further enhanced through integration with a wide range of datasets, from a coin’s everyday results for the last ten years, enabling the application to capture the predicted stock prices.

Reports