One unique property of time series is that the temporal relations are largely preserved after downsampling into two sub-sequences. By taking advantage of this property, we propose a novel neural network architecture that conducts sample convolution and interaction for temporal modeling and forecasting, named SCINet. Specifically, SCINet is a recursive downsample-convolve-interact architecture. In each layer, we use multiple convolutional filters to extract distinct yet valuable temporal features from the downsampled sub-sequences or features. By combining these rich features aggregated from multiple resolutions, SCINet effectively models time series with complex temporal dynamics. Experimental results show that SCINet achieves significant forecasting accuracy improvements over both existing convolutional models and Transformer-based solutions across various real-world time series forecasting datasets. Our codes and data are available at https://github.com/cure-lab/SCINet.
翻译:时间序列的一个独特属性是,时间关系在被降格为两个子序列后基本上得以保存。我们利用这一属性,提出了一个新的神经网络结构,进行样本变异和互动,用于时间建模和预测。具体地说,SCINet是一个循环的下标-相交互动结构。在每一层,我们使用多个革命过滤器,从下标次序列或特征中提取不同但宝贵的时间特征。通过将这些来自多个分辨率的丰富特征结合起来,SCINet有效地模拟时间序列,同时进行复杂的时间动态。实验结果显示,SCINet在各种实时时间预测数据集中,对现有革命模型和基于变形器的解决方案都实现了显著的预测准确性改进。我们的代码和数据可在https://github.com/cure-lab/SCINet上查阅。