Prediction of The Diabetics Using The Deep Learning Based Artificial Memory Vector Network

An rise in blood sugar or a lack of insulin causes diabetes, one of the most prevalent, severe, and expensive illnesses worldwide today. If diabetes is not detected in time, it might have serious consequences for the health of the patient. The rise of diabetes has become a major problem for health and government agencies. Taking care of people's health without sacrificing their comfort is crucial, and preventing illness is a top responsibility. In this investigation, an accurate diagnosis and prognosis for the onset of diabetes mellitus were achieved by using an optimization and deep learning based approach. In this first step, error is eliminated from the dataset's input data by replicative derivative error normalization. Using a multivariate salp swarm optimization (MSSO) strategy, unique features may be retrieved from an artificial bound co-occurrence matrix (ABCM). Finally, artificial memory vector network (AMVN) classifier based on deep learning is used to classify/predict diabetes risk as part of the classification technique. In this research, genuine data from the University of California's website on Indians with diabetes were used for experimental purposes. When compared to other current mechanisms, the findings demonstrate the superior performance of the suggested technique in illness diagnosis, which reaches 99.8% accuracy in this investigation.