DBLoss: Decomposition-based Loss Function for Time Series Forecasting

2025-10-26

AI Summary

Traditional mean squared error (MSE) loss struggles to accurately model trend and seasonality components in time series forecasting. To address this, we propose DBLoss, a novel decomposition-based loss function. Its core innovation lies in the first integration of sequence decomposition directly into loss design: exponential moving average (EMA) is applied synchronously to both predicted and ground-truth sequences to extract trend and seasonal components, and separate losses are computed for each; these are then jointly optimized via learnable weights. DBLoss requires no architectural modifications and can be seamlessly integrated into any deep learning forecasting framework as a drop-in replacement. Extensive experiments across multiple real-world datasets demonstrate that DBLoss consistently improves forecasting accuracy of state-of-the-art models—including Transformer and Informer—validating its effectiveness, broad applicability, and strong generalization capability.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Search and Optimization: Learning to Search
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Time series forecasting holds significant value in various domains such as economics, traffic, energy, and AIOps, as accurate predictions facilitate informed decision-making. However, the existing Mean Squared Error (MSE) loss function sometimes fails to accurately capture the seasonality or trend within the forecasting horizon, even when decomposition modules are used in the forward propagation to model the trend and seasonality separately. To address these challenges, we propose a simple yet effective Decomposition-Based Loss function called DBLoss. This method uses exponential moving averages to decompose the time series into seasonal and trend components within the forecasting horizon, and then calculates the loss for each of these components separately, followed by weighting them. As a general loss function, DBLoss can be combined with any deep learning forecasting model. Extensive experiments demonstrate that DBLoss significantly improves the performance of state-of-the-art models across diverse real-world datasets and provides a new perspective on the design of time series loss functions.

Problem

• Addresses MSE limitations in capturing seasonality and trend
• Proposes decomposition-based loss function for time series forecasting
• Improves forecasting accuracy across diverse real-world datasets

Innovation

• Decomposes time series into seasonal and trend components
• Calculates separate losses for each component with weighting
• Combines with any deep learning forecasting model