Predicting the Severity of Diabetes Using ECLAT Algorithm in Data Mining

Diabetes is a long-term disease that damages the various parts of the human body. According to the World Health Organization (WHO) report, there was a deficiency of 12.9 million medical care workers estimated in 2035. Various Artificial Intelligence (AI) and Machine Learning (ML) classifiers were used to anticipate and diagnose diabetes. In the case of unsupervised ML classifications, data mining acts an important role in the diagnosis and prediction of the disease. Selecting legitimate classifiers clearly expands the correctness and adeptness of the proposed system. Public awareness of the disease is very poor in India. Deficient healthcare facilities lead to the growth of the disease in families. Apriori, FP growth, and ECLAT are the different types of association rule mining algorithms for the diagnosis and prediction of diabetes. Equivalence Class clustering and bottom-up LAttice Traversal (ECLAT) algorithm is used for the prediction of the severity of the diabetes in the proposed paper. The proposed work will provide a new platform for analyzing the data set for new patients and submitting an accurate prediction. Pregnancy frequency, diastolic blood pressure, diabetes pedigree function, and class distribution outcome, etc. are the parameters considered for the prediction of the severity range of diabetes. This paper aims to develop a model for an Intelligent Diabetes Prediction system using the ECLAT algorithm and reduce medical misdiagnoses by providing proper interpretation and bringing down treatment costs.