Diabetes Analysis Based on Naive Bayes Model

Abstract

As a matter of fact, diabetes has been one of the most disturbing diseases in human beings’ daily life in recent years. With this in mind, the method to predict diabetes has been conducted through artificial intelligence technology including machine learning as well as big data analysis. On this basis, this study is going to discuss the possibility of predicting diabetes via one of the widely used machine learning scenarios, i.e., Naive Bayes models. According to the analysis, the result shows that although the accuracy is improved after parameter changing, the value is 75.61%, still below 80%. In other words, it indicates that the features offered nowadays are still abundant enough to realize high accuracy predictions. Despite the low accuracy, the result still can be used and applied in many fields, such as the pre-processing of another model’s building. Overall, these results shed light on guiding further exploration of diabetes analysis.