Memory-Gated Recurrent Networks

Thirty-Fifth AAAI Conference on Artificial Intelligence (AAAI-21)

Posted: 1 Mar 2021

See all articles by Yaquan Zhang

Yaquan Zhang

affiliation not provided to SSRN

Qi Wu

City University of Hong Kong, School of Data Science

Nanbo Peng

JD Digits

Min Dai

The Hong Kong Polytechnic University

Jing Zhang

University of Nottingham

Hu Wang

JD Digits

Date Written: December 2020

Abstract

The essence of multivariate sequential learning is all about how to extract dependencies in data. These data sets, such as hourly medical records in intensive care units and multi-frequency phonetic time series, often time exhibit not only strong serial dependencies in the individual components (the "marginal" memory) but also non-negligible memories in the cross-sectional dependencies (the "joint" memory). Because of the multivariate complexity in the evolution of the joint distribution that underlies the data generating process, we take a data-driven approach and construct a novel recurrent network architecture, termed Memory-Gated Recurrent Networks (mGRN), with gates explicitly regulating two distinct types of memories: the marginal memory and the joint memory. Through a combination of comprehensive simulation studies and empirical experiments on a range of public datasets, we show that our proposed mGRN architecture consistently outperforms state-of-the-art architectures targeting multivariate time series.

Suggested Citation

Zhang, Yaquan and Wu, Qi and Peng, Nanbo and Dai, Min and Zhang, Jing and Wang, Hu, Memory-Gated Recurrent Networks (December 2020). Thirty-Fifth AAAI Conference on Artificial Intelligence (AAAI-21) , Available at SSRN: https://ssrn.com/abstract=3755716

Yaquan Zhang

affiliation not provided to SSRN

Qi Wu (Contact Author)

City University of Hong Kong, School of Data Science ( email )

83 Tat Chee Avenue
Kowloon
Hong Kong

Nanbo Peng

JD Digits

Min Dai

The Hong Kong Polytechnic University ( email )

Jing Zhang

University of Nottingham ( email )

Hu Wang

JD Digits ( email )

China

Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
203
PlumX Metrics