Prediction of Gestational Diabetes Mellitus under Cascade and Ensemble Learning Algorithm

Area of Science:

• Medical Informatics
• Computational Biology
• Machine Learning in Healthcare

Background:

• Gestational diabetes mellitus (GDM) poses significant risks to both mother and fetus.
• Early prediction of GDM is highly sought after to mitigate pregnancy complications.

Purpose of the Study:

• To develop and evaluate a machine learning model for accurate and efficient prediction of GDM risk.
• To identify key predictive features for GDM using feature engineering techniques.

Main Methods:

• Data preprocessing including outlier handling with mean values.
• Feature selection using the IV value method, reducing 83 features to 40.
• Ensemble of multiple machine learning models: Logistics regression, Lasso-Logistics, GBDT, Xgboost, LightGBM, and Catboost.

Main Results:

• The proposed ensemble model achieved an accuracy of 80.3% and a recall rate of 79.3%.
• The model demonstrated superior performance compared to previous methods in accuracy, precision, recall, and F1-score.
• The prediction model completed in a rapid 2.53 seconds, meeting engineering requirements.

Conclusions:

• The developed machine learning model offers a promising approach for early GDM risk prediction.
• The study highlights the effectiveness of ensemble methods and feature engineering in disease prediction.
• This research provides valuable insights for applying machine learning in clinical disease risk assessment.