Intelligent Prediabetes and Type-2 Diabetes Prediction from the Genomic Data using an Optimal Framework

Abstract

Objective: Diabetes is a chronic illness with a high prevalence that has a negative impact on people's quality of life and a high death rate in both industrialized and developing nations. The patient’s diabetes type 2 prediction is the most critical study in medical research. Several prediction models exist to predict type 2 diabetes. However, the relevant result was not satisfied because of poor quality. The gene data contains Nan features, which maximize the complexity of predicting type 2 diabetes. The demerits resulted in low prediction and performance scores.

Material and Methods: The proposed work aims to develop a novel chimp-based functional link neural approach (CbFLNA) to predict type 2 diabetes. The proposed methodology is pre-processing; feature selection, classification, and gene expression have been performed. Initially, the genomic database was imported, the data were pre-processed, and the meaning features were extracted. Then, predict the type 2 diabetes of the patients and classify the conditions.

Results: The performance was measured and the model attained 99.7% accuracy, 99.7% precision, 99.7% recall, 99.7% F-score, 0.002 error rate and execution time of 74.1547s.

Conclusion: The presented model attained a high exactness score in the prediction.