Papers for

forecasting software engineers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

SACQ improves long-term time series forecasting with structured decoding

SACQ: Structured Decoding with Memory-Conditioned Refinement for Long-Horizon Forecasting

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.

Thu 8 OctMachine Learning
The gist
Long-term forecasting of time-based data often struggles because current models treat all future predictions as linked in a way that hides how specific past moments relate to each future event. The authors introduce SACQ, a new method that first makes a rough forecast and then improves each future point by focusing on the most relevant past information. This approach makes predictions more accurate and resistant to noisy or corrupted data. It works well with existing forecasting models and keeps computational cost low.
Open → 2610.11170v1

Visual bridge helps control time series data for better forecasting

Image Synthesis as an Intermediate for Controllable Time Series Generation

Abstract: Semantic-driven time-series generation offers a promising way to improve downstream learning in few-shot forecasting, but directly generating numerical sequences from language often fails to preserve the intended temporal structure. We propose VisualBridge, which uses time-series plots as a visual intermediate to bridge high-level temporal semantics and numerical sequences. An MLLM first converts plotted series into structured semantic representations, enabling explicit control over temporal properties such as trend, seasonality, and volatility. We then learn a semantic editing policy with downstream forecasting rewards, allowing the generation process to favor temporal patterns that are beneficial for the target task. The resulting sequences are further modeled by a temporal VAE to produce consistent multivariate augmentations. Experiments on standard public forecasting benchmarks demonstrate that VisualBridge improves few-shot forecasting over conventional augmentation methods, with ablations validating the roles of visual semantic grounding, learned semantic control, and VAE-based generation.

Sun 4 OctArtificial Intelligence
The gist
Time series data, like stock prices or weather patterns, can be hard to generate correctly from just descriptions because the timing and patterns matter a lot. The authors created VisualBridge, which first turns these descriptions into pictures of the data, then uses those images to better capture and control key patterns like trends or seasonal changes. This makes the generated data more useful for predicting future events even when only a little real data is available. Their tests show this method helps forecasting tools learn better than usual data augmentation tricks.
Open → 2610.05211v1