An Advanced Learning Based Diabetes Mellitus Prediction Using KNN

Diabetes Mellitus (DM) is one of the chronic diseases that variations in blood sugars may cause. If it is not identified in the early stages, this may lead to severe issues over health. Diabetes may cause many side effects to the human body. Thus, detecting and diagnosing in the early stages is very important. This paper uses the advanced learning model to detect diabetes mellitus using K-Nearest Neighbours (KNN) combined with other hybrid techniques. The quality and relevance of the input were improved by using various data preprocessing techniques and feature selection methods to solve missing value issues. The KNN in this scenario performed better because of its intelligibility and efficiency in classifying the DM to obtain predictive results. The performance of the proposed approach was analyzed using confusion matrix metrics, which show actual values and predicted values. The results show that the proposed approach outperforms the results based on the given dataset.