Enhancing Diabetes Prediction Using Advanced Machine Learning Techniques

Abstract

Diabetes has been an issue over the past few decades, where it keeps on having an upward trend throughout each year. One of the biggest contributing factor is caused by the possibility that people can remain undiagnosed during their early stages of diabetes. Therefore, this research is done to help develop a machine learning good enough to diagnose early-stage diabetes. The development of this machine learning will use the Pima Indian Diabetes Dataset, since another dataset is unable to be obtained. The methods to achieve a great model for identifying diabetes will be divided to 5 parts which are data preparation, exploratory data analysis, feature engineering, build the model, and evaluate it. There are various kinds of techniques on finishing each steps mentioned that are tried, however only some techniques support the model well.