SWoTTeD: An Extension of Tensor Decomposition to Temporal Phenotyping

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Sebia, Hana, Guyet, Thomas, Audureau, Etienne
Formato: Preprint
Publicado: 2023
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866909186890137600
author Sebia, Hana
Guyet, Thomas
Audureau, Etienne
author_facet Sebia, Hana
Guyet, Thomas
Audureau, Etienne
contents Tensor decomposition has recently been gaining attention in the machine learning community for the analysis of individual traces, such as Electronic Health Records (EHR). However, this task becomes significantly more difficult when the data follows complex temporal patterns. This paper introduces the notion of a temporal phenotype as an arrangement of features over time and it proposes SWoTTeD (Sliding Window for Temporal Tensor Decomposition), a novel method to discover hidden temporal patterns. SWoTTeD integrates several constraints and regularizations to enhance the interpretability of the extracted phenotypes. We validate our proposal using both synthetic and real-world datasets, and we present an original usecase using data from the Greater Paris University Hospital. The results show that SWoTTeD achieves at least as accurate reconstruction as recent state-of-the-art tensor decomposition models, and extracts temporal phenotypes that are meaningful for clinicians.
format Preprint
id arxiv_https___arxiv_org_abs_2310_01201
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle SWoTTeD: An Extension of Tensor Decomposition to Temporal Phenotyping
Sebia, Hana
Guyet, Thomas
Audureau, Etienne
Machine Learning
Tensor decomposition has recently been gaining attention in the machine learning community for the analysis of individual traces, such as Electronic Health Records (EHR). However, this task becomes significantly more difficult when the data follows complex temporal patterns. This paper introduces the notion of a temporal phenotype as an arrangement of features over time and it proposes SWoTTeD (Sliding Window for Temporal Tensor Decomposition), a novel method to discover hidden temporal patterns. SWoTTeD integrates several constraints and regularizations to enhance the interpretability of the extracted phenotypes. We validate our proposal using both synthetic and real-world datasets, and we present an original usecase using data from the Greater Paris University Hospital. The results show that SWoTTeD achieves at least as accurate reconstruction as recent state-of-the-art tensor decomposition models, and extracts temporal phenotypes that are meaningful for clinicians.
title SWoTTeD: An Extension of Tensor Decomposition to Temporal Phenotyping
topic Machine Learning
url https://arxiv.org/abs/2310.01201