WaveGNN: Integrating Graph Neural Networks and Transformers for Decay-Aware Classification of Irregular Clinical Time-Series

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hajisafi, Arash, Siampou, Maria Despoina, Azarijoo, Bita, Xiong, Zhen, Shahabi, Cyrus
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917149529866240
author Hajisafi, Arash
Siampou, Maria Despoina
Azarijoo, Bita
Xiong, Zhen
Shahabi, Cyrus
author_facet Hajisafi, Arash
Siampou, Maria Despoina
Azarijoo, Bita
Xiong, Zhen
Shahabi, Cyrus
contents Clinical time series are often irregularly sampled, with varying sensor frequencies, missing observations, and misaligned timestamps. Prior approaches typically address these irregularities by interpolating data into regular sequences, thereby introducing bias, or by generating inconsistent and uninterpretable relationships across sensor measurements, complicating the accurate learning of both intra-series and inter-series dependencies. We introduce WaveGNN, a model that operates directly on irregular multivariate time series without interpolation or conversion to a regular representation. WaveGNN combines a decay-aware Transformer to capture intra-series dynamics with a sample-specific graph neural network that models both short-term and long-term inter-sensor relationships. Therefore, it generates a single, sparse, and interpretable graph per sample. Across multiple benchmark datasets (P12, P19, MIMIC-III, and PAM), WaveGNN delivers consistently strong performance, whereas other state-of-the-art baselines tend to perform well on some datasets or tasks but poorly on others. While WaveGNN does not necessarily surpass every method in every case, its consistency and robustness across diverse settings set it apart. Moreover, the learned graphs align well with known physiological structures, enhancing interpretability and supporting clinical decision-making.
format Preprint
id arxiv_https___arxiv_org_abs_2412_10621
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle WaveGNN: Integrating Graph Neural Networks and Transformers for Decay-Aware Classification of Irregular Clinical Time-Series
Hajisafi, Arash
Siampou, Maria Despoina
Azarijoo, Bita
Xiong, Zhen
Shahabi, Cyrus
Machine Learning
Artificial Intelligence
Clinical time series are often irregularly sampled, with varying sensor frequencies, missing observations, and misaligned timestamps. Prior approaches typically address these irregularities by interpolating data into regular sequences, thereby introducing bias, or by generating inconsistent and uninterpretable relationships across sensor measurements, complicating the accurate learning of both intra-series and inter-series dependencies. We introduce WaveGNN, a model that operates directly on irregular multivariate time series without interpolation or conversion to a regular representation. WaveGNN combines a decay-aware Transformer to capture intra-series dynamics with a sample-specific graph neural network that models both short-term and long-term inter-sensor relationships. Therefore, it generates a single, sparse, and interpretable graph per sample. Across multiple benchmark datasets (P12, P19, MIMIC-III, and PAM), WaveGNN delivers consistently strong performance, whereas other state-of-the-art baselines tend to perform well on some datasets or tasks but poorly on others. While WaveGNN does not necessarily surpass every method in every case, its consistency and robustness across diverse settings set it apart. Moreover, the learned graphs align well with known physiological structures, enhancing interpretability and supporting clinical decision-making.
title WaveGNN: Integrating Graph Neural Networks and Transformers for Decay-Aware Classification of Irregular Clinical Time-Series
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2412.10621