Improving efficiency in Diabetes Detection using data mining techniques

Abstract

Diabetes mellitus is seen as a disease which affects highly social, human and financial newlineexpenses for a nation. Simultaneously, it is necessary to reduce the rate of newlinecommonness along with this the confusions needed to be resolved regarding diabetes. newlineLearning is the ability to improve with experience, remember past decisions and newlineachievements in order to make better decisions in similar situations in the future. newlineMachine learning is a discipline of artificial intelligence. One of the most important newlinereasons for creating machine learning systems is that in many areas experience is newlineinsufficient, and the codification of the knowledge that describes it is limited, newlinefragmented and, therefore, incomplete. Artificial intelligence has made significant newlineadvances in fields such as education, agriculture and healthcare, where it can detect newlineand treat diseases like cancer and diabetes long before traditional methods. The first newlinephase of this study will present a comparative neural network analysis and will newlineimplement a neural network classifier optimized by Firefly algorithm to predict newlinediabetes diagnosis based on factors mentioned in patients from all diabetes data from newlinePIMA Indian database. newlineThe second phase of this work will develop a multi-scale convolutional neural newlinenetwork (MCNN) model for the early diagnosis of diabetes mellitus. The detection newlinetechnique is based on a proposed tanning model by analyzing data from diabetic and newlinenon-diabetic patients from the PIMA Indian database. To get accurate training data newlinefor high-grade diabetes, we finally recorded parameters such as diastolic blood newlinepressure, body mass index (BMI), triceps skin thickness, number of pregnancies, newlineheredity and age. Based on MCNNs operating at different resolutions, the proposed newlinearchitecture avoids the traditional steps of manually extracting resources, extracting newlineresources and simultaneously classifying them into a single neural data network. The newlineproposed approach offers better classification results than the conventional method for newlinea more reliable diagnosis of dia