Machine Learning Algorithms for Side Effects Prediction: A Comparative Analysis

Drug's side effects prediction is considered as one of the most important topics in the field of pharmacy and drug discovery, as pharmaceutical factories dedicate significant resources to this goal. This has led researchers to contemplate finding more cost-effective and efficient methods. Machine learning has found ample opportunities to contribute to this field, yielding promising results. Our research proposes a comparison between five machine learning classifiers i.e., Random Forest, Decision Tree, K-Nearest Neighbor, Naive Bayes and Support Vector Machine, and deep learning model, Multi-Layer Perceptron, to predict the occurrence of stomach pain based on certain drug properties. Our research shows that the Random Forest model demonstrated superiority in all cases, achieving an accuracy up to 92%. Through this research, we can gain an indepth understanding of the effectiveness of each model and its performance under the same conditions and on the same dataset.