FV-MgNet: Fully Connected V-cycle MgNet for Interpretable Time Series Forecasting
This work addresses a difficult forecasting problem for time series analysis, but it is incremental as it adapts an existing CNN model from image classification to time series.
The authors tackled long-term time series forecasting by proposing FV-MgNet, a fully connected V-cycle version of MgNet, which achieved better results with less memory usage and faster inference speed compared to state-of-the-art models on popular datasets.
By investigating iterative methods for a constrained linear model, we propose a new class of fully connected V-cycle MgNet for long-term time series forecasting, which is one of the most difficult tasks in forecasting. MgNet is a CNN model that was proposed for image classification based on the multigrid (MG) methods for solving discretized partial differential equations (PDEs). We replace the convolutional operations with fully connected operations in the existing MgNet and then apply them to forecasting problems. Motivated by the V-cycle structure in MG, we further propose the FV-MgNet, a V-cycle version of the fully connected MgNet, to extract features hierarchically. By evaluating the performance of FV-MgNet on popular data sets and comparing it with state-of-the-art models, we show that the FV-MgNet achieves better results with less memory usage and faster inference speed. In addition, we develop ablation experiments to demonstrate that the structure of FV-MgNet is the best choice among the many variants.