Abstract
Long-term time series forecasting (LTSF) models predominantly employ patch-based encoders terminated by a flatten readout head that maps the entire encoded historical memory to all future steps through a single shared projection. This implicit coupling of future positions obscures position-specific historical-to-future alignment and amplifies sensitivity to corrupted inputs and extreme supervision noise. We present SACQ, a plug-in structured prediction head that replaces flatten readout while keeping the encoder unchanged. SACQ adopts a two-stage decoding pipeline: it first establishes a coarse patch-grid forecast scaffold, then refines each future position through cross-attention over historical memory and merges the attention-derived correction with the coarse scaffold via a learned per-patch gate. To stabilize optimization under long horizons and noisy labels, we further propose a batch-adaptive scaled log-cosh loss that automatically calibrates robustness to the current residual scale, suppressing outlier gradients while preserving MSE-like sensitivity for typical errors. SACQ attains top-tier test MSE/MAE across PatchTST, DLinear, and patch-Mamba backbones with only modest incremental overhead in parameters and latency. Under inference-time input corruption and training-set label-noise stress tests, SACQ substantially outperforms flatten readouts, with ablation studies validating each architectural component.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Cheng, G., Xu, Z., Jing, C., & Hou, J. (2026). SACQ: Structured Decoding with Memory-Conditioned Refinement for Long-Horizon Forecasting. https://omanscience.com/en/articles/sacq-structured-decoding-with-memory-conditioned-refinement-for-long-horizon-forecasting
MLA 9
Cheng, Guo, et al. "SACQ: Structured Decoding with Memory-Conditioned Refinement for Long-Horizon Forecasting." https://omanscience.com/en/articles/sacq-structured-decoding-with-memory-conditioned-refinement-for-long-horizon-forecasting.
Chicago (author–date)
Cheng, Guo, Zhengzhuo Xu, Chenchen Jing, and Jingyi Hou. 2026. "SACQ: Structured Decoding with Memory-Conditioned Refinement for Long-Horizon Forecasting." https://omanscience.com/en/articles/sacq-structured-decoding-with-memory-conditioned-refinement-for-long-horizon-forecasting.
Harvard
Cheng, G., Xu, Z., Jing, C. and Hou, J. (2026) 'SACQ: Structured Decoding with Memory-Conditioned Refinement for Long-Horizon Forecasting', Available at: https://omanscience.com/en/articles/sacq-structured-decoding-with-memory-conditioned-refinement-for-long-horizon-forecasting.
Vancouver
Cheng G, Xu Z, Jing C, Hou J. SACQ: Structured Decoding with Memory-Conditioned Refinement for Long-Horizon Forecasting. https://omanscience.com/en/articles/sacq-structured-decoding-with-memory-conditioned-refinement-for-long-horizon-forecasting
IEEE
G. Cheng, Z. Xu, C. Jing, and J. Hou, "SACQ: Structured Decoding with Memory-Conditioned Refinement for Long-Horizon Forecasting," https://omanscience.com/en/articles/sacq-structured-decoding-with-memory-conditioned-refinement-for-long-horizon-forecasting.