Hi community,<\/p>\n
I recently published a research-based synthetic Chronic Kidney Disease (CKD) dataset on Kaggle after spending several weeks studying clinical guidelines and epidemiological literature.<\/p>\n
The motivation came from a common challenge I encountered: many publicly available CKD datasets contain only a few hundred patient records and a limited number of clinical variables, making them less suitable for building and evaluating modern machine learning models.<\/p>\n
Dataset Highlights<\/p>\n
\u2022 200,000 synthetic patient records<\/p>\n
\u2022 82 clinically meaningful features<\/p>\n
\u2022 Research-informed design using published clinical guidelines and epidemiological evidence<\/p>\n
\u2022 Covers demographics, lifestyle, medical history, vital signs, kidney biomarkers, medications, frailty, healthcare utilization, and clinical outcomes<\/p>\n
\u2022 Includes CKD stage, kidney failure risk, dialysis requirement, and hospitalization risk<\/p>\n
\u2022 Machine learning and healthcare analytics ready<\/p>\n
The dataset is completely synthetic and contains no real patient information. It was created for educational purposes, machine learning experiments, healthcare analytics, and research.<\/p>\n
I’d really appreciate feedback from the community.<\/p>\n
Some questions I’d love your thoughts on:<\/p>\n
\u2022 Are there any important CKD-related variables you think are missing?<\/p>\n
\u2022 What types of ML or analytics projects would you build with this dataset?<\/p>\n
\u2022 What would you improve in a future version?<\/p>\n
Kaggle Dataset:<\/p>\n
https:\/\/www.kaggle.com\/datasets\/mohankrishnathalla\/chronic-kidney-disease-risk-dataset-2026<\/a><\/p>\n Thanks for taking the time to check it out. I’m happy to answer questions about the design process or discuss future improvements.<\/p>\n<\/div>\n submitted by \/u\/Mohan137 <\/a>
[link]<\/a><\/span> [comments]<\/a><\/span><\/p>
