In many operational time-series forecasting applications, such as crowd demand forecasting, the risk associated with under-prediction is significantly higher than that of over-prediction. Accurate prediction of rare demand spikes plays a crucial role in downstream tasks. However, most time-series forecasters are trained with symmetric objectives (e.g., MSE, MAE) and evaluated primarily on aggregate error, which can obscure failures in extreme values and peak timing predictions. We introduce Asymmetric Peak-Aware Loss (APAL), a simple, model-agnostic objective that:
- Heavily penalizes under-predictions.
- Increases the training weight of peak regions within each forecast window.
We further propose a peak-critical evaluation protocol that complements MAE/MSE with channel-wise tail error (Top-10% and Top-1%) and peak metrics (precision, recall, F1 under timing tolerance, and peak timing error). We evaluate APAL on long-horizon multivariate forecasting across five state-of-the-art backbones, focusing on pedestrian demand forecasting using (i) a production-ready subset of the City of Melbourne pedestrian hourly count dataset and (ii) a beach visitor count dataset. The generality of the loss function for time-series forecasting is tested on additional benchmarks. Across peak-critical datasets and settings, APAL improves tail accuracy and peak prediction quality while exposing a controllable trade-off with aggregate error, making it a practical solution when peak prediction failures are the dominant operational concern.
Blogger's Review: The introduction of APAL marks a significant innovation in the time-series forecasting domain, particularly in addressing demand peaks. By effectively enhancing prediction accuracy and reliability, its design principles are worth noting for other fields, especially in constructing loss functions to handle asymmetric risks. Focusing on peak regions allows APAL to provide greater practicality in real-world applications.