Comparative analysis using classification methods versus early stage diabetes

TL;DR

This study compares Discriminant Analysis and Logistic Regression to predict early stage diabetes using a 2020 UC Irvine dataset, highlighting differences in their effectiveness.

Contribution

It provides a comparative analysis of two classification methods for early diabetes prediction, emphasizing the practical differences and insights for medical data analysis.

Findings

Logistic Regression was more frequently used in disease classification studies.

Significant differences were observed between the two classification methods.

The analysis offers valuable guidance for selecting classification techniques in medical diagnostics.

Abstract

In this research work, a comparative analysis was carried out using classification methods such as: Discriminant Analysis and Logistic Regression to subsequently predict whether a person may have the presence of early stage diabetes. For this purpose, use was made of a database of the UC IRVINE platform of the year 2020 where specific variables that influence diabetes were used for a better result. Likewise in terms of methodology, the corresponding analysis was performed for each of the 3 classification methods and then take them to a comparative table and analyze the results obtained. Finally we can add that the majority of the studies carried out applying the classification methods to the diseases can be clearly seen that there is a certain attachment and more use of the logistic regression classification method, on the other hand, in the results we could see significant differences in terms of the 2 classification methods that were applied, which was valuable information for later drawing final conclusions.