Antoska-Knights, Vesna and Gavriloska, Elena (2024) Machine Learning Techniques for Modelling and Predicting the Influence of Kefir in a Low-Protein Diet on Kidney Function. Medical Research Archives.
Full text not available from this repository.Abstract
Introduction: This study aims to investigate the effects of a low-protein diet supplemented with kefir on protein catabolism and kidney function in stable chronic kidney disease patients. By employing advanced machine learning techniques, this research will explore the potential impact of kefir as a probiotic fermented dairy product on kidney function within the recommended intake. The study seeks to understand whether kefir supplementation in a low-protein diet can help maintain kidney function and identify any potential benefits or limitations.
Methods: The study employed a dataset comprising kidney health indicators and kefir intake records collected from 150 randomly selected patients in stage G1 to G5 during one year. Data preprocessing was performed to ensure data quality and feature relevance. Subsequently, a range of machine learning algorithms, including decision trees, random forests, and neural networks, but also and Stochastic Gradient Boosting and XgBoost model were implemented to model and predict the impact of kefir on kidney function.
The clinical data: age, sex, blood pressure (systolic and diastolic), BMI, albumin, bacteria, level of blood glucose, hemoglobin, creatinine, urea, UNAPCR, protein intake and kefir, GFR, MDRD, proturija, the existence of hypertension, diabetes mellitus, coronary artery disease, stage of CKD at the beginning and CKD stage after 12 months, binary output (patient stay in a same stage or is reduced kidney function).
Results: The analysis results revealed promising predictive capabilities of the machine learning models, demonstrating associations between kefir consumption and kidney function. Binary output indicates the patient stayed in the same CKD stage using low-protein diet where source of protein is kefir.
Conclusion: This research underscores the value of machine learning techniques in modeling and predicting the impact of kefir on kidney function. By shedding light on potential associations, this study paves the way for further investigations into the role of kefir in kidney health and sets a precedent for future studies in this area.
Keywords: Kefir, low protein diet, chronic kidney disease, machine learning technique, Modeling and the Prediction
| Item Type: | Article |
|---|---|
| Subjects: | Scientific Fields (Frascati) > Natural sciences > Mathematics |
| Divisions: | Faculty of Technology and Technical Sciences |
| Depositing User: | Mr Jordan Martinovski |
| Date Deposited: | 01 Nov 2024 13:29 |
| Last Modified: | 01 Nov 2024 13:29 |
| URI: | https://eprints.uklo.edu.mk/id/eprint/10426 |
Actions (login required)
 |
View Item |