Related Experiment Video
Updated: Aug 26, 2025
12:18
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
7.6K
Gradient boosting decision tree becomes more reliable than logistic regression in predicting probability for diabetes
Hiroe Seto1,2, Asuka Oyama3, Shuji Kitora1
1Health Care Division, Health and Counseling Center, Osaka University, Osaka, 560-0043, Japan.
Scientific Reports
|October 11, 2022
Summary
Gradient boosting decision tree (GBDT) models demonstrate superior reliability over logistic regression (LR) for diabetes prediction using big data. Machine learning (ML) offers a powerful tool for developing accurate diabetes prediction models to aid lifestyle improvements and prevention strategies.
Area of Science:
- Computational epidemiology
- Health informatics
- Machine learning in healthcare
Background:
- Accurate diabetes prediction is crucial for public health and preventative strategies.
- Machine learning (ML) offers potential for developing robust predictive models using large datasets.
- Comparing different ML algorithms is essential to determine the most reliable methods for clinical application.
Purpose of the Study:
- To evaluate and compare the reliability of gradient boosting decision tree (GBDT) and logistic regression (LR) models for diabetes prediction.
- To assess the performance of these models using big data from Japanese health checkup records.
- To analyze the impact of sample size on model reliability.
Main Methods:
- Utilized a large dataset of 277,651 participants from the Kokuho-database in Osaka, Japan.
- Developed diabetes prediction models using LightGBM (a GBDT implementation) and logistic regression (LR) with 16 health checkup predictors.
- Assessed model reliability using expected calibration error (ECE), negative log-likelihood (Logloss), and area under the curve (AUC).
Main Results:
- LightGBM achieved a lower ECE (0.0018) and Logloss (0.167) and a higher AUC (0.844) compared to LR (ECE: 0.0048, Logloss: 0.172, AUC: 0.826).
- The reliability of the LightGBM model surpassed LR when the training sample size exceeded a specific threshold.
- 15,900 participants were newly diagnosed with diabetes within three years, validating the prediction task.
Conclusions:
- Gradient boosting decision tree (GBDT) models, specifically LightGBM, offer superior reliability and accuracy for diabetes prediction compared to logistic regression when using big data.
- Machine learning (ML) holds significant promise for developing highly reliable diabetes prediction tools.
- These ML-driven models can serve as valuable instruments for promoting healthier lifestyles and preventing diabetes.
More Related Videos
07:13Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
180
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
8.3K
Related Concept Videos
01:19Survival Tree
133
Survival trees are a non-parametric method used in survival analysis to model the relationship between a set of covariates and the time until an event of interest occurs, often referred to as the "time-to-event" or "survival time." This method is particularly useful when dealing with censored data, where the event has not occurred for some individuals by the end of the study period, or when the exact time of the event is unknown.
Building a Survival Tree
Constructing a...
Building a Survival Tree
Constructing a...
133
01:15Receiver Operating Characteristic Plot
313
A ROC (Receiver Operating Characteristic) plot is a graphical tool used to assess the performance of a binary classification model by illustrating the trade-off between sensitivity (true positive rate) and specificity (false positive rate). By plotting sensitivity against 1 - specificity across various threshold settings, the ROC curve shows how well the model distinguishes between classes, with a curve closer to the top-left corner indicating a more accurate model. The area under the ROC curve...
313
01:25Statistical Methods for Analyzing Epidemiological Data
484
Epidemiological data primarily involves information on specific populations' occurrence, distribution, and determinants of health and diseases. This data is crucial for understanding disease patterns and impacts, aiding public health decision-making and disease prevention strategies. The analysis of epidemiological data employs various statistical methods to interpret health-related data effectively. Here are some commonly used methods:
484
01:22Diabetes Mellitus: Type 2 and Gestational
2.7K
Type 2 diabetes, characterized by insulin resistance, arises when the insulin receptors on cells lose responsiveness to insulin, diminishing the cell's capacity to take up glucose, resulting in elevated blood glucose levels. To receive a diagnosis of Type 2 diabetes, a series of blood glucose tests are necessary to assess whether the blood glucose falls within normal parameters. If the result is out of the normal range, a patient may be diagnosed as prediabetic or diabetic, depending on the...
2.7K
01:24Bootstrapping
660
The term "bootstrap" originated in the 19th century as a metaphor for self-improvement or achieving something independently, without external assistance. This concept extends to statistical bootstrapping, a self-contained method for estimating population parameters through resampling, even though it can be computationally intensive. Developed by the American statistician Dr. Bradley Efron in 1979, bootstrapping provides a robust way to perform inference when the original sample size is...
660
01:11Residuals and Least-Squares Property
7.8K
The vertical distance between the actual value of y and the estimated value of y. In other words, it measures the vertical distance between the actual data point and the predicted point on the line
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
7.8K