Optimizing Healthcare Outcomes with a Medical Recommendation System Based on Machine Learning

Abstract

Health care is extremely reliant on technology in a digital age, and this plays an important part in the fight against many diseases.Despite technological advancements, Misdiagnosis continue to pose a significant global health challenge. For avoiding these risks, This thesis is to explore how Machine Learning algorithms can be used in the disease diagnosis by building an ML-based Medical Recommendation System that significantly enhances the accuracy of disease diagnosis .Additionally, Even in an age where technology has become far more advanced and information is readily available to many people, a lot of people still follow traditional long-winded methods for seeking medical attention which can be time-consuming. This thesis also proposes a method for providing complete and detailed treatment recommendations these recommendations include descriptions of diseases, precautions, medications, exercise routines, and dietary suggestions tailored to the patient’s needs using machine learning. Utilizing two datasets of patient symptoms to train and test the models, we found that for dataset 1 the MultinomialNB model performed best at 97.36\%,followed by SVC at 94.44\%. Regarding dataset 2, the DNN model performed best at 84.19\%. This study implies that ML and DL algorithms could decreasing misdiagnosis, and improving patient care. This thesis illustrates a strong framework for application of advanced technologies in healthcare highlighting their transformative impact and substantial benefits, ultimately optimizing resource utilization for doctors and enhancing care and information for patients.