Machine learning based Comprehensive Study for Stock Market Prediction of Pharmaceutical Industry Index on Covid 19

Abstract

This research examines how COVID-19 vaccinations impact the accuracy of machine-learning models in forecasting the Tehran Stock Exchange's Pharmaceutical Companies Index. The study analyses daily vaccination and stock data during the pandemic using statistical and linear regression models. Results reveal a negative correlation between vaccinations and the stock index. Two regression models were developed, one with vaccination data and one without. Although both models fit the training data well, the latter performed significantly better on the test set with lower errors. This suggests that vaccination data does not enhance the predictive ability of the regression model for the stock index during the pandemic. In fact, excluding vaccination data leads to better predictive performance. Therefore, accelerating vaccination programs could aid in the stock market recovery. However, avoiding vaccination data as an input feature for machine learning models forecasting this pharmaceutical stock index is advisable.