Título:
Towards foundation auto-encoders for time-series anomaly detection.
Otros títulos:
Coordinador:
Director:
Compilador:
Autor:
García González, Gastón
Casas, Pedro
Martínez, Emilio
Fernández, Alicia
Casas, Pedro
Martínez, Emilio
Fernández, Alicia
Tutor:
Tipo de documento:
Ponencia
Editor:
Palabras clave:
Time-Series Data
Anomaly Detection
VAE
Foundation Models
Anomaly Detection
VAE
Foundation Models
Descriptores:
Año de publicación:
2024
Contenido:
Resumen:
We investigate a novel approach to time-series modeling, inspired by the successes of large pretrained foundation models. We introduce FAE (Foundation Auto-Encoders), a foundation generative-AI model for anomaly detection in time-series data, based on Variational Auto-Encoders (VAEs). By foundation, we mean a model pretrained on massive amounts of time-series data which can learn complex temporal patterns useful for accurate modeling, forecasting, and detection of anomalies on previously unseen datasets. FAE leverages VAEs and Dilated Convolutional Neural Networks (DCNNs) to build a generic model for univariate time-series modeling, which could eventually perform properly in out-of-the-box, zero-shot anomaly detection applications. We introduce the main concepts of FAE, and present preliminary results in different multidimensional time-series datasets from various domains, including a real dataset from an operational mobile ISP, and the well known KDD 2021 Anomaly Detection dataset.
Descripción:
metadata.articulos.dc.description.uri:
Editorial:
MILETS
EN:
The 10th Mining and Learning from Time Series Workshop : From Classical Methods to LLMs (MILETS 2024), Aug 25th, 2024 - KDD 2024, Barcelona, Spain, pp. 1-9.
Financiadores:
Este trabajo ha sido financiado parcialmente por FWF Austrian Science Fund, Project I-6653 GRAPHS4SEC, the Austrian FFG ICT- of-the-Future project DynAISEC – Adaptive AI/ML for Dynamic Cybersecurity Systems – ID 887504 y por el Proyecto uruguayo CSIC con referencia CSIC-I+D-22520220100371UD Generalization and Domain Adaptation in Time-Series Anomaly Detection.
Citación:
García González, G., Casas, P., Martínez, E. y otros. Towards foundation auto-encoders for time-series anomaly detection [en línea]. EN: The 10th Mining and Learning from Time Series Workshop : From Classical Methods to LLMs (MILETS 2024), Barcelona, Spain, Aug 25th 2024, pp. 1-9.
Citación:
metadata.articulos.cc.license.name:
ISBN:
e-ISBN:
ISSN:
ISMN:
Otros identificadores:
Cobertura geográfica:
Cobertura temporal:
Departamento académico:
Procesamiento de Señales
Grupo de investigación:
Tratamiento de Imagenes
Licencia:
Licencia Creative Commons Atribución - No Comercial - Sin Derivadas (CC - By-NC-ND 4.0)
Colecciones:
| Ficheros | Descripción | Tamaño | Formato | ||
|---|---|---|---|---|---|
| GCMF24.pdf | Versión definitiva | 2.84 MB | Adobe PDF |