AliO: Output Alignment Matters in Long-Term Time Series Forecasing

arXiv cs.AI Papers
time-series-forecasting long-term-prediction alignment deep-learning evaluation-metric

Summary

本文提出 AliO（Align Outputs）方法，通过在时域和频域减少相同时间戳上预测输出的差异来提升长期时间序列预测模型的输出对齐性，并引入新指标 TAM（Time Alignment Metric）量化对齐程度，在保持或提升预测性能的同时最高提升 58.2% 的对齐度。

arXiv:2610.11213v1 Announce Type: new Abstract: Long-term Time Series Forecasting (LTSF) tasks, which leverage the current data sequence as input to predict the future sequence, have become increasingly crucial in real-world applications such as weather forecasting and planning of electricity consumption. However, state-of-the-art LTSF models often fail to achieve prediction output alignment for the same timestamps across lagged input sequences. Instead, these models exhibit low output alignment, resulting in fluctuation in prediction outputs for the same timestamps, undermining the model's reliability. To address this, we propose AliO (Align Outputs), a novel approach designed to improve the output alignment of LTSF models by reducing the discrepancies between prediction outputs for the same timestamps in both the time and frequency domains. To measure output alignment, we introduce a new metric, TAM (Time Alignment Metric), which quantifies the alignment between prediction outputs, whereas existing metrics such as MSE only capture the distance between prediction outputs and ground truths. Experimental results show that AliO effectively improves the output alignment, i.e., up to 58.2% in TAM, while maintaining or enhancing the forecasting performance (up to 27.5%). This improved output alignment increases the reliability of the LTSF models, making them more applicable in real-world scenarios.
Original Article
View Cached Full Text

Cached at: 10/09/26, 06:26 AM 

# AliO: Output Alignment Matters in Long-Term Time Series Forecasing
Source: [https://arxiv.org/abs/2610.11213](https://arxiv.org/abs/2610.11213)
[View PDF](https://arxiv.org/pdf/2610.11213)[HTML \(experimental\)](https://arxiv.org/html/2610.11213v1)

> Abstract:Long\-term Time Series Forecasting \(LTSF\) tasks, which leverage the current data sequence as input to predict the future sequence, have become increasingly crucial in real\-world applications such as weather forecasting and planning of electricity consumption\. However, state\-of\-the\-art LTSF models often fail to achieve prediction output alignment for the same timestamps across lagged input sequences\. Instead, these models exhibit low output alignment, resulting in fluctuation in prediction outputs for the same timestamps, undermining the model's reliability\. To address this, we propose AliO \(Align Outputs\), a novel approach designed to improve the output alignment of LTSF models by reducing the discrepancies between prediction outputs for the same timestamps in both the time and frequency domains\. To measure output alignment, we introduce a new metric, TAM \(Time Alignment Metric\), which quantifies the alignment between prediction outputs, whereas existing metrics such as MSE only capture the distance between prediction outputs and ground truths\. Experimental results show that AliO effectively improves the output alignment, i\.e\., up to 58\.2% in TAM, while maintaining or enhancing the forecasting performance \(up to 27\.5%\)\. This improved output alignment increases the reliability of the LTSF models, making them more applicable in real\-world scenarios\.

## Submission history

From: Kwangryeol Park \[[view email](https://arxiv.org/show-email/5ecb03c8/2610.11213)\] **\[v1\]**Thu, 8 Oct 2026 04:12:35 UTC \(4,186 KB\)

Similar Articles

Alignment Forecasting: Predicting Misalignment From Training Data

arXiv cs.CL

The paper introduces Alignment Forecasting, a task for predicting misalignment failure modes from training data before a model is fine-tuned, along with AlignmentForecastBench, a benchmark of over 5,000 forecasting questions covering 17 models, 32 datasets, and 16 failure modes. Their LLM-based forecasting scaffold outperforms baselines and helps filter problematic training examples to produce more aligned models in most evaluated cases.

To Align or Not To Align: Check Your COMPASS Before You Train

Hugging Face Daily Papers

论文提出 COMPASS（COMPatibility ASsessment Score），通过测量局部邻域重叠与范式上限，在对齐或合并之前低成本预测模型对的兼容性，在 72 个跨模态对齐实验和 392 个任务算术合并中与下游成功度高度相关（Pearson 相关分别达 0.91 和 0.70）。