Carleton University - School of Computer Science Honours Project
Fall 2019
Predicting stock prices and trends in the video game market
Michael Lambucki
SCS Honours Project Image
ABSTRACT
Predicting stock prices and market trends have always been a difficult problem to solve. The purpose of this project is to address this issue through the development of a program that attempts to estimate the stock prices of a select few video game companies. Using the Keras API and TensorFlow, the program creates a Recurrent Neural Network with a built-in Long Short Term Memory cell to forecast the company’s closing price for the following day. Furthermore, the program’s Bidirectional Long Short Term Memory network analyzes the reviews of recent game releases in order to determine the sentiment of the reviews. Through the application of both these systems, the program examines and predicts future stock market trends of the selected video game companies and displays them in graph form on a Graphical User Interface.