Why the study?
Exploring important features of diabetes through data mining analytical methods can help predict and prevent diabetes.
Comparison
Improved XGBoost algorithm with combined numerical and text features
Authors
Loading...
May support ML diabetes prediction tools; leaves open prospective validation before clinical use.
An improved XGBoost algorithm using combined numerical and text features achieved 80.2% accuracy in predicting diabetes.
Li et al. (2020) studied this question.