ImputeINR: Time Series Imputation via Implicit Neural Representations for Disease Diagnosis with Missing Data

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Li, Mengxuan, Liu, Ke, Guo, Jialong, Bu, Jiajun, Wang, Hongwei, Wang, Haishuai
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912379379384320
author Li, Mengxuan
Liu, Ke
Guo, Jialong
Bu, Jiajun
Wang, Hongwei
Wang, Haishuai
author_facet Li, Mengxuan
Liu, Ke
Guo, Jialong
Bu, Jiajun
Wang, Hongwei
Wang, Haishuai
contents Healthcare data frequently contain a substantial proportion of missing values, necessitating effective time series imputation to support downstream disease diagnosis tasks. However, existing imputation methods focus on discrete data points and are unable to effectively model sparse data, resulting in particularly poor performance for imputing substantial missing values. In this paper, we propose a novel approach, ImputeINR, for time series imputation by employing implicit neural representations (INR) to learn continuous functions for time series. ImputeINR leverages the merits of INR in that the continuous functions are not coupled to sampling frequency and have infinite sampling frequency, allowing ImputeINR to generate fine-grained imputations even on extremely sparse observed values. Extensive experiments conducted on eight datasets with five ratios of masked values show the superior imputation performance of ImputeINR, especially for high missing ratios in time series data. Furthermore, we validate that applying ImputeINR to impute missing values in healthcare data enhances the performance of downstream disease diagnosis tasks. Codes are available.
format Preprint
id arxiv_https___arxiv_org_abs_2505_10856
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ImputeINR: Time Series Imputation via Implicit Neural Representations for Disease Diagnosis with Missing Data
Li, Mengxuan
Liu, Ke
Guo, Jialong
Bu, Jiajun
Wang, Hongwei
Wang, Haishuai
Machine Learning
Artificial Intelligence
Healthcare data frequently contain a substantial proportion of missing values, necessitating effective time series imputation to support downstream disease diagnosis tasks. However, existing imputation methods focus on discrete data points and are unable to effectively model sparse data, resulting in particularly poor performance for imputing substantial missing values. In this paper, we propose a novel approach, ImputeINR, for time series imputation by employing implicit neural representations (INR) to learn continuous functions for time series. ImputeINR leverages the merits of INR in that the continuous functions are not coupled to sampling frequency and have infinite sampling frequency, allowing ImputeINR to generate fine-grained imputations even on extremely sparse observed values. Extensive experiments conducted on eight datasets with five ratios of masked values show the superior imputation performance of ImputeINR, especially for high missing ratios in time series data. Furthermore, we validate that applying ImputeINR to impute missing values in healthcare data enhances the performance of downstream disease diagnosis tasks. Codes are available.
title ImputeINR: Time Series Imputation via Implicit Neural Representations for Disease Diagnosis with Missing Data
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2505.10856