自动进化生成时间序列预测代码，减少人工干预
SEA-TS: Self-Evolving Agent for Autonomous Code Generation of Time Series Forecasting Algorithms
- 用自适应优势评分引导搜索，提升算法探索效率
- 通过代码执行反馈持续优化提示词，实现迭代改进
- 适合需要自动化建模的工业场景，尤其数据少时
准确的时间序列预测支撑众多领域的决策，但传统机器学习开发常受限于数据稀缺、分布漂移和手动调优收益递减。本文提出自进化代理框架SEATS，通过迭代自进化循环自主生成、验证与优化时间序列预测算法代码。设计包含三项机制：(1) 基于统计归一化优势得分的度量优势蒙特卡洛树搜索（MA-MCTS），替代固定奖励以更好指导搜索；(2) 代码执行后进行评审并动态优化提示词，使成功解法中的修正模式被编码用于后续迭代；(3) 全局可调节推理机制，将每个评估节点与全局最优/最差解对比，实现跨轨迹知识迁移。一个MAP-Elites档案维护架构多样性。在四个数据集和两个指标下，SEATS在八次比较中胜出七次，优于TimeMixer、Timer和SEMixer等强基线。
原文摘要 · Abstract (English)
Accurate time series forecasting underpins decision-making in many domains, yetconventional ML development often faces data scarcity, distribution shift, anddiminishing returns from manual iteration. We propose Self-Evolving Agent forTime Series Algorithms (SEATS), a framework that autonomously generates, val-idates, and optimizes forecasting algorithm code through an iterative self-evolutionloop. Our design combines three mechanisms: (1) Metric-Advantage MCTS(MA-MCTS), which replaces fixed rewards with a statistically normalized advan-tage score for search guidance, (2) code review with running prompt refinement,so every successfully executed solution is reviewed and the running prompt encodescorrective patterns for later iterations, and (3) global steerable reasoning, whichcompares each evaluated node to global best- and worst-performing solutions forcross-trajectory transfer. A MAP-Elites archive maintains architectural diversity.Across four datasets and two metrics, SEATS wins seven of eight comparisonsagainst strong baselines TimeMixer, Timer, and SEMixer
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。