Researchers have developed a new dataset, DiaData, by integrating 15 existing sources to create a comprehensive collection of over 149 million continuous glucose monitoring measurements from 2510 individuals with Type 1 diabetes. This dataset, which includes demographic and heart rate information, aims to address the scarcity of large-scale data for diabetes research. A related paper investigates the effectiveness of age-specialized models versus a general population model for predicting hypoglycemia, finding that a combined model performs comparably or better, though children's data may benefit from age-specific training. AI
影响 Provides a large, integrated dataset for diabetes research and offers insights into model generalization for hypoglycemia prediction.
排序理由 The cluster contains two arXiv papers presenting a new dataset and research on its application.
- arXiv
- Beyza Cinar
- continuous glucose monitoring
- DiaData
- hypoglycemia
- machine learning
- Type 1 diabetes
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
