A Machine Learning Approach: Enhancing the Predictive Performance of Pharmaceutical Stock Price Movement during COVID

Predicting stock price movement direction is a challenging problem influenced by different factors and capricious events. The conventional prediction machine learning models heavily rely on the internal financial features, especially history. However, there are many outside-of-company features that deeply interact with companies’ performance, during COVID period. In this study, we selected 9 vaccine companies collected their relevant over past 20 months. We added handcrafted external information, including COVID-related statistics company-specific progress information. implemented, evaluated, compared several models, Multilayer Perceptron Neural Networks logistic regression decision trees boosting bagging algorithms. results suggest application of feature engineering data mining techniques can effectively enhance performance predicting show help to increase model accuracy 7.3% AUROC 6.5% average. Further exploration showed selection tree gradient, algorithm achieved 70% in 66% accuracy.