Summary of Health-llm: Large Language Models For Health Prediction Via Wearable Sensor Data, by Yubin Kim et al.
Health-LLM: Large Language Models for Health Prediction via Wearable Sensor Data
by Yubin Kim, Xuhai Xu, Daniel McDuff, Cynthia Breazeal, Hae Won Park
First submitted to arxiv on: 12 Jan 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper explores the capacity of large language models (LLMs) to make accurate inferences about consumer health based on contextual information and physiological data. The study fine-tunes 12 state-of-the-art LLMs with prompting techniques on four public health datasets, covering 10 consumer health prediction tasks in mental health, activity, metabolic, and sleep assessment. Notably, the fine-tuned model, HealthAlpaca, achieves comparable performance to larger models like GPT-3.5, GPT-4, and Gemini-Pro, outperforming them in 8 out of 10 tasks. The results highlight the effectiveness of context enhancement strategies, with a maximum improvement of 23.8% observed when combining user context, health knowledge, and temporal information. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how well large language models can understand consumer health based on different kinds of data. They tested 12 of these models to see how good they were at making predictions about mental health, physical activity, metabolic rate, and sleep quality. The best model, called HealthAlpaca, did almost as well as some much bigger models. The study also found that adding extra information about the person’s health knowledge can help the model make better predictions. |
Keywords
* Artificial intelligence * Gemini * Gpt * Prompting
