The Prediction of Tesla Stock Price During COVID-19 Utilizing Machine Learning Methods

Abstract

Stock forecasting is an act of predicting the future trend of the stock price based on the fluctuation of the stock price over a period of time. Due to the economic lockdown caused by the COVID-19 pandemic, it is challenging to provide accurate forecasts of price movements to ensure the reliability of investments. This study used Long Short-Term Memory (LSTM), Gate Recurrent Unit (GRU), Random Forest and K-Nearest Neighbor (KNN) algorithms to predict the stock price of Tesla in the special period of the epidemic. In the training data set, the fluctuation situation during the epidemic period was added as the basis for prediction, and as expected, the stock trend of Tesla in the special period was more accurately predicted. According to the results, LSTM and GRU models have high accuracy. They can provide investors with more reliable information about stock prices, and their R-squared values reach nearly 0.95, while the rest two models had poor performances.