基于混合模糊隶属函数的特征变换用于心血管疾病预测
A novel hybrid fuzzy membership function-based feature transformation for cardiovascular disease prediction
研究者提出一种结合高斯函数与 S 型函数的混合模糊隶属函数特征变换方法,并与梯度提升决策树(GBDT)及 Bagging Fuzzy‑GBDT 结合用于心血管疾病预测。该模型在 UCI 心脏病数据集上准确率达 98%,在 Kaggle Framingham 数据集上准确率为 84.23%,10 折交叉验证显示其泛化稳定。结果提示该方法对可靠、可解释的心血管病预测具潜力,但需进一步外部与临床验证。
We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply.
Abstract
Cardiovascular disease portrays a critical health issue due to its huge numbers of morbidity and mortality rate. Stable, accurate prediction and diagnosis are important for early intervention, monitoring, and treatment. The area of Artificial Intelligence (AI) and Internet of Medical Things (IoMT) significantly enhanced the healthcare sector, providing early detection of diseases, diagnosis, monitoring, and personalized care. Despite the advancement in AI-based diagnosis system existing methods often struggle with handling uncertainty, complexity, discreteness in datasets which leads to model overfitting, poses a significant challenge to develop a generalized model. This study proposes a novel hybrid fuzzy membership function-based feature transformation technique. The proposed technique combines gaussian function to capture localized precision in high-confidence region and sigmoid functions ensure smooth transition in area with greater certainty. Subsequently, the proposed hybrid fuzzy membership function (F-MF) integrated with gradient boosting decision tree (GBDT) and bagging Fuzzy-GBDT for heart disease prediction. The proposed approaches reduce overfitting, adapt to complex, nonlinear decision boundaries and improve predictability of individual features, resulting in improved model generalization and overall prediction performance. The validation of the proposed methodology is carried out on two publicly available benchmark datasets: (1) UCI Heart Disease dataset and (2) a publicly available Kaggle version of the Framingham dataset for cardiovascular risk prediction. The proposed model on Dataset 1 achieved accuracy of 98%, precision of 98%, recall of 98%, F1 score of 99% and AU-ROC (Area Under- Receiver Operating Characteristics) score 0.99. The model achieved an accuracy of 84.23%, precision of 88.86%, recall of 70.49%, F1 score of 78.62% and AU-ROC of 0.91 on Dataset 2. Furthermore, 10-fold cross-validation confirms the model’s generalizability, with minimum variance in performance across all folds. These results highlight the potential of hybrid fuzzy membership functions with GBDT models for reliable, interpretable CVD prediction subject to further external and clinical validation.
Funding
The authors declare no financial support was received for the research.
Author information
Author notes
These authors contributed equally this work: Abhay Kumar Pathak and Mrityunjay Chaubey.
Authors and Affiliations
Department of Computer Science, Institute of Science, Banaras Hindu University, Varanasi, 221005, India
Abhay Kumar Pathak, Mrityunjay Chaubey & Marisha
School of Computer Science, UPES University, Dehradun, India
Mrityunjay Chaubey
Department of Computer Science and Coordinator- Centre for Interdisciplinary Mathematical Sciences, Institute of Science, Banaras Hindu University, Varanasi, 221005, India
Manjari Gupta
Authors
- Abhay Kumar Pathak
- Mrityunjay Chaubey
- Marisha
- Manjari Gupta
Corresponding authors
Correspondence to Mrityunjay Chaubey or Manjari Gupta.
Ethics declarations
Competing interests
The authors declare no competing interests.
Ethical approval and consent for data
The cardiovascular datasets employed in this study constitutes publicly available data. Consequently, ethical approval was not required, and the research adhered to relevant data usage and ethical guidelines.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Pathak, A.K., Chaubey, M., Marisha et al. A novel hybrid fuzzy membership function-based feature transformation for cardiovascular disease prediction. Sci Rep (2026). https://doi.org/10.1038/s41598-026-76878-y
Received: 01 May 2026
Accepted: 15 October 2026
Published: 19 November 2026
DOI: https://doi.org/10.1038/s41598-026-76878-y
Keywords
来源:Nature 健康护理主题 · nature.com