Summary of Xtsformer: Cross-temporal-scale Transformer For Irregular-time Event Prediction in Clinical Applications, by Tingsong Xiao et al.
XTSFormer: Cross-Temporal-Scale Transformer for Irregular-Time Event Prediction in Clinical Applications
by Tingsong Xiao, Zelin Xu, Wenchong He, Zhengkun Xiao, Yupu Zhang, Zibo Liu, Shigang Chen, My T. Thai, Jiang Bian, Parisa Rashidi, Zhe Jiang
First submitted to arxiv on: 3 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
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 proposes a novel deep learning model called XTSFormer to predict clinical events from electronic health records (EHRs). The authors focus on modeling de facto care pathways, which are common step-by-step plans for treatment or care. Existing neural temporal point processes (TPPs) methods struggle to capture the multi-scale nature of event interactions, which is prevalent in many real-world clinical applications. To address this issue, XTSFormer incorporates two key components: a Feature-based Cycle-aware Time Positional Encoding (FCPE) and a hierarchical multi-scale temporal attention mechanism. The FCPE component captures the cyclical nature of time, while the attention mechanism determines different temporal scales through a bottom-up clustering approach. Experimental results on several real-world EHR datasets demonstrate that XTSFormer outperforms multiple baseline methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us better understand how to use computers to analyze medical data and make predictions about what might happen in the future. The goal is to improve patient safety by identifying potential problems before they occur. The authors created a new way of analyzing this kind of data, which includes recognizing patterns and cycles that are important for understanding how people’s health changes over time. |
Keywords
* Artificial intelligence * Attention * Clustering * Deep learning * Positional encoding
