为多通道时间序列设计感知损失，提升各通道预测精度。
CP Loss: Channel-wise Perceptual Loss for Time Series Forecasting
- 为每个通道学习独立的感知空间，捕捉其特有动态。
- 通过可学习滤波器分解信号，构建多尺度表示。
- 适合关注通道差异的时序预测任务，如金融、气象建模。
多通道时间序列在众多应用中广泛存在，其不同通道间具有显著异质性。然而，现有预测模型通常采用通道无关的损失函数（如MSE），对所有通道使用统一度量，难以捕捉各通道特有的动态特征，例如剧烈波动或趋势突变。为此，本文提出一种通道感知损失（CP Loss）。核心思想是为每个通道学习一个适配其特性的独特感知空间，并在该空间内计算损失。具体而言，设计了一个可学习的通道自适应滤波器，将原始信号分解为解耦的多尺度表示，构成感知空间基础。关键在于，该滤波器与主预测模型联合优化，确保所学感知空间明确服务于预测任务。最终，在各感知空间内计算损失以优化模型。代码已公开于 https://github.com/zyh16143998882/CP_Loss。
原文摘要 · Abstract (English)
Multi-channel time-series data, prevalent across diverse applications, is characterized by significant heterogeneity in its different channels. However, existing forecasting models are typically guided by channel-agnostic loss functions like MSE, which apply a uniform metric across all channels. This often leads to fail to capture channel-specific dynamics such as sharp fluctuations or trend shifts. To address this, we propose a Channel-wise Perceptual Loss (CP Loss). Its core idea is to learn a unique perceptual space for each channel that is adapted to its characteristics, and to compute the loss within this space. Specifically, we first design a learnable channel-wise filter that decomposes the raw signal into disentangled multi-scale representations, which form the basis of our perceptual space. Crucially, the filter is optimized jointly with the main forecasting model, ensuring that the learned perceptual space is explicitly oriented towards the prediction task. Finally, losses are calculated within these perception spaces to optimize the model. Code is available at https://github.com/zyh16143998882/CP_Loss.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。