A Hybrid Model for Prediction of Diabetes Using Machine Learning Classification Algorithms and Random Projection

The medical industry has risen quickly to be particularly interested in the concept of machine learning. Clinical sets of information used in research predictions and studies aid in preventative care by offering effective interventions and monitoring. One of the diseases that is spreading fastest in the world is diabetes, which needs to be monitored constantly. We investigate various machine learning techniques that will aid in the early diagnosis of this disease to verify this. In this study, a reliable diabetes prediction system using Random Projection feature reduction has been developed. the hybrid approach of random projection with machine learning classification algorithms can be used for disease prediction. In the field of medical informatics, high-dimensional data, such as gene expression data or imaging data, is often used to predict diseases or diagnose conditions. The use of random projection in combination with machine learning classification algorithms can help reduce the dimensionality of the data and improve the accuracy of disease prediction. Three steps make up the proposed framework's overall process: (1) Pre-processing (2) Dimensionality reduction using Random Projection (3) Prediction using Hybrid Classifier. For categorization, a number of Machine Learning (ML) methods are employed, including Bayesian Net, Ada Boost, Logit Boost, Decision Table and Hoeffding Tree. This study investigated whether RP is best suited for use with machine learning methods to increase illness prediction accuracy. In this case, the dataset from the repository of UCI is used for testing in order to assess the system's performance and compare it to assessment criteria such as classification accuracy, True Positive ROC Curve, PRC Curve and Precision. The results of this study suggest that the proposed hybrid method is an effective and reliable approach for improving disease prediction accuracy, potentially aiding in earlier diagnosis and better healthcare outcomes.