Enhanced Diabetes Prediction Using Optimization and Clustering Techniques with Secure Cloud Storage

One of the chronic diseases is diabetes that is rapidly increasing globally, impacting 422 million people. Because diabetes has a protracted asymptomatic period, detecting the disease early is essential to producing clinically meaningful outcomes. Around half of all diabetics go undetected because of this asymptomatic phase. Correct identification of typical and less typical indications seen at different phases of the illness is essential for effective detection. Generally speaking, diabetes is characterised by high blood sugar and perhaps insufficient insulin synthesis. The proportion of diabetic patients worldwide has dramatically increased recently. Reducing the overall number of diabetes cases requires taking this issue seriously. Numerous research groups have investigated various data mining systems for analysing health data, including patient datasets with diabetes. Using parametric analysis, data mining (DM) can synthesise useful knowledge from health data. In order to assess medical data pertaining to diabetes diagnosis, this study presents the Iterative Cuckoo Search Optimisation clustering algorithm, which is based on adaptive gradient density. The Deep VGG Net classifier is also utilised to distinguish between data that has diabetes and data that does not. The prime contribution of the research is the use for Refined ELGamal security algorithms to ensure cloud security while mining patient disease results.