Improving Diagnostic Accuracy: A Deep Dive into Random Forest Optimization for Clinical Data

The healthcare industry produces a large amount of patient data, which makes it possible to use a variety of analytical techniques on a large dataset. A predictive system that can recognize several diseases at once has been created by utilizing machine learning techniques. Grid Search and Randomized Search strategies are used to further enhance the model that was developed using the Random Forest algorithm. This contrasts with widely used systems that frequently display a numerous illness prediction system without improving the model, producing erroneous and subpar results. Reduced accuracy and erroneous results have serious consequences that could endanger the health of the patient. Our system’s present focus is on two different diseases: Diabetes and Parkinson’s disease. Especially, the architecture of the system makes it possible to easily incorporate other illnesses in later stages. Users are required to input various disease-specific parameters, after which the system promptly generates output indicating the presence or absence of the specified condition. This research work carried out demonstrates the disease prediction system developed using Optimized Machine learning algorithms such as Decision Tree classifier, Random forest classifier, Support Vector Machine (SVM) and Logistic Regression along with Optimization techniques Grid Search and Randomized Search. The primary objective of the model is to enhance its recall capabilities through the incorporation of advanced methodologies and optimization techniques. The paper also presents comparative study of the results of the above algorithms used.