FEATURE SELECTION FOR INSULIN RESISTANCE USING RANDOM FORESTS BASED APPROACH

Abstract

Type-2 diabetes mellitus (T2DM) is a significant problem considering that it is anticipated to reach over 693 million people by 2045. T2DM is the serious situation of insulin resistance, recognition, and quantification of insulin resistance calls for a specific blood examination which is made complex, lengthy, and most notably intrusive, making it not possible for regular day-to-day tasks of a human. With the advancement of current Artificial Intelligence approaches, the identification of insulin resistance might be executed without clinical procedures. In this job, insulin resistance is determined based on Machine Learning methods utilizing non-invasive strategies. Nineteen parameters are used for recognition of insulin resistance; such as age, sex, waist size, height, and so on as well as a mix of these specifications. Experiments are performed on the CALERIE dataset to determine the factors that impact insulin resistance. Each result of the function option technique is modelled with the help of a Random Forest Classifier. The suggested technique is validated by making use of a Stratified cross-validation test. Outcomes reveal that utilizing Logistic regression, Naïve Bayes, LDA, and also Random forests Classifier for recognition of insulin resistance, precision as much as 0.843 with AUC scores of 0.84 using Naïve Bayes classifier. [32] The major benefit of the suggested approach is that a person may forecast the insulin resistance and hence future probabilities of diabetic issues may be checked daily utilizing non-clinical methods. While the same is not virtually possible with clinical procedures.