MULTIPLE DISEASE PREDICTION AND DRUG RECOMMENDATION SYSTEM USING MACHINE LEARNING

COMPUTER ENGINEERING

The increasing prevalence of chronic diseases and the exponential growth of healthcare data have created a strong need for intelligent, data-driven diagnostic solutions. Traditional medical diagnosis often requires expert interpretation, repeated clinical examinations, and significant time. These limitations encourage the use of machine learning-based decision-support systems that can efficiently analyze patient attributes and predict potential diseases in real-time. This research work proposes a unified Multiple Disease Prediction and Drug Recommendation System capable of predicting three major diseases—Diabetes, Heart Disease, and Parkinson’s Disease—using supervised machine learning algorithms. The system integrates essential stages such as data preprocessing, feature standardization, model training, evaluation, and deployment through an interactive web interface developed using Streamlit. The machine learning models used include Logistic Regression for diabetes prediction, Random Forest for heart disease prediction, and Support Vector Machine (SVM) for Parkinson’s disease detection. A rule-based drug recommendation module is incorporated to suggest suitable medications based on commonly prescribed drugs for each predicted condition. Publicly available UCI datasets were used for training and evaluation, and results show high accuracy for all models, with SVM achieving 94% for Parkinson’s, Random Forest achieving 88% for heart disease, and Logistic Regression yielding 82% accuracy for diabetes. The proposed system is designed to assist healthcare providers and individuals in early detection, reducing human error, and offering timely treatment guidance. Overall, this work demonstrates the effectiveness of machine learning in healthcare analytics and highlights its potential to enhance preventive medicine and personalized treatment.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.