AI Based Drug Recommendation And Diseases Prediction With Patient Assistance System

Abstract

The proposed system uses machine learning to improve disease prediction and drug suggestion. Combining Random Forest, Decision Tree and Logistic Regression with a web interface built on Django offers real-time, precise medical diagnoses and customized drug recommendations. The system utilises two datasets—Disease.csv for disease prediction and Drug.csv for drug recommendation, which include structured data. These are preprocessed, encoded, and trained using ensemble learning techniques to enhance model performance. The system dramatically reduces the workload of healthcare professionals and minimizes the risk of human error. Drug recommendations are further validated through manual checks to ensure patient safety. Various machine learning algorithms are studied and implemented to showcase the accuracy and reliability of the system. The performance of the models is summarized as follows: the Random Forest model achieved a precision of 84%, recall of 83%, F1-score of 83.5%, and accuracy of 91%; the Decision Tree model achieved a precision of 86%, recall of 87%, F1-score of 86.5%, and accuracy of 88%; while the Logistic Regression model achieved a precision of 62%, recall of 60%, F 1 -score of 61%, and accuracy of 61%. These metrics demonstrate the effectiveness of each model in handling medical data for disease detection and drug suggestion. Support Vector Machine (SVM) will be used to compare model performance for benchmarking.

Ähnliche Arbeiten