TY - GEN
T1 - EnhanceNet: Plugin Neural Networks for Enhancing Correlated Time Series Forecasting
AU - Cirstea, Razvan Gabriel
AU - Kieu, Tung
AU - Guo, Chenjuan
AU - Yang, Bin
AU - Pan, Sinno Jialin
N1 - Publisher Copyright:
© 2021 IEEE.
PY - 2021/4
Y1 - 2021/4
N2 - Correlated time series forecasting plays an essential role in many cyber-physical systems, where entities interact with each other over time. To enable accurate forecasting, it is essential to capture both the temporal dynamics and the correlations among different entities. To capture the former, two popular types of models, recurrent neural networks (RNNs) and temporal convolution networks (TCNs), are employed. To capture the latter, a graph is constructed to reflect certain relationships among entities and then graph convolution (GC) is applied upon the graph to capture the correlations among the entities. The state-of-the-art forecasting accuracy is achieved by models that combine RNNs or TCNs with GC. However, they neither capture distinct temporal dynamics that exist among different entities nor consider the entity correlations that evolve across time.In this paper, rather than proposing yet another new end-to-end forecasting model, we aim at providing a framework to enhance existing forecasting models, where we propose generic plugins that can be easily integrated into existing solutions to solve the two challenges and thus further enhance their accuracy. Specifically, we propose two plugin neural networks that are able to better capture distinct temporal dynamics for different entities and dynamic entity correlations across time, so that forecasting accuracy is improved while model parameters to be learned are reduced. Experimental results on three real-world correlated time series data sets demonstrate that the proposed framework with the two plugin networks is able to achieve the above goals.
AB - Correlated time series forecasting plays an essential role in many cyber-physical systems, where entities interact with each other over time. To enable accurate forecasting, it is essential to capture both the temporal dynamics and the correlations among different entities. To capture the former, two popular types of models, recurrent neural networks (RNNs) and temporal convolution networks (TCNs), are employed. To capture the latter, a graph is constructed to reflect certain relationships among entities and then graph convolution (GC) is applied upon the graph to capture the correlations among the entities. The state-of-the-art forecasting accuracy is achieved by models that combine RNNs or TCNs with GC. However, they neither capture distinct temporal dynamics that exist among different entities nor consider the entity correlations that evolve across time.In this paper, rather than proposing yet another new end-to-end forecasting model, we aim at providing a framework to enhance existing forecasting models, where we propose generic plugins that can be easily integrated into existing solutions to solve the two challenges and thus further enhance their accuracy. Specifically, we propose two plugin neural networks that are able to better capture distinct temporal dynamics for different entities and dynamic entity correlations across time, so that forecasting accuracy is improved while model parameters to be learned are reduced. Experimental results on three real-world correlated time series data sets demonstrate that the proposed framework with the two plugin networks is able to achieve the above goals.
KW - Dynamic weights
KW - Neural networks
KW - Time series forecasting
UR - http://www.scopus.com/inward/record.url?scp=85097264905&partnerID=8YFLogxK
U2 - 10.1109/ICDE51399.2021.00153
DO - 10.1109/ICDE51399.2021.00153
M3 - Article in proceeding
AN - SCOPUS:85097264905
T3 - Proceedings - International Conference on Data Engineering
SP - 1739
EP - 1750
BT - Proceedings of the 37th IEEE International Conference on Data Engineering, ICDE 2021
PB - IEEE (Institute of Electrical and Electronics Engineers)
T2 - 37th IEEE International Conference on Data Engineering, ICDE 2021
Y2 - 19 April 2021 through 22 April 2021
ER -