An overview of diabetes diagnosis methods on the Pima Indian dataset

Abstract

In recent years, data mining and machine learning methods in the medical field have received much attention and have optimized many complex issues in the medical field. One of the problems facing researchers is the appropriate dataset, and the suitable dataset on which different methods of data mining and machine learning can be applied is rarely found. One of the most reliable and appropriate datasets in the field of diabetes diagnosis is the Indian Survey Database. In this article, we have tried to review the methods that have been implemented in recent years using machine learning classification algorithms on this data set and compare these methods in terms of evaluation criteria and feature selection methods. After comparing these methods, it was found that models that used feature selection methods were more accurate than other approaches.

Keywords

• diabetes
• machine learning
• data mining algorithms
• detection accuracy
• pima indian dataset