Model identification for ARMA time series through convolutional neural networks

We use convolutional neural networks for model identification in ARMA time series models, where our networks are trained on synthetic data with known ground truths. Comparing the performance of these networks with traditional likelihood-based methods, in particular the Akaike and Bayesian Informatio...

Full description

Saved in:
Bibliographic Details
Published in:Decision Support Systems 2021-07, Vol.146, p.113544, Article 113544
Main Authors: Tang, Wai Hoh, Röllin, Adrian
Format: Article
Language:English
Subjects:
AIC
Artificial neural networks
Autoregressive moving average time series (ARMA)
BIC
Convolutional neural networks
Model selection
Neural networks
Statistical inference
Time series
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We use convolutional neural networks for model identification in ARMA time series models, where our networks are trained on synthetic data with known ground truths. Comparing the performance of these networks with traditional likelihood-based methods, in particular the Akaike and Bayesian Information Criteria, we are able to show that when it comes to statistical inference on ARMA orders, neural networks can significantly outperform likelihood-based methods in terms of accuracy and, by orders of magnitude, in terms of speed. We also observe improvements in terms of time series forecasting. Our approach shows the feasibility of using artificial neural networks for statistical inference in situations where classical likelihood-based methods are difficult or costly to implement.
ISSN:0167-9236
1873-5797
DOI:10.1016/j.dss.2021.113544