AlphaCast: A Human Wisdom-LLM Intelligence Co-Reasoning Framework for Interactive Time Series Forecasting

TL;DR

AlphaCast reframes time-series forecasting as an interactive, co-reasoned process between human intelligence and LLMs to address the limitations of static, one-shot predictors in dynamic environments. It introduces a two-stage framework: prediction preparation (feature extraction, knowledge base, contextual repository, and case library) and generative reasoning with reflective optimization (including a reflector for meta-reasoning). Empirical results across short- and long-horizon datasets show consistent improvements in predictive accuracy ($MSE$/$MAE$) over fourteen baselines, with strong performance on high-volatility and seasonally driven series. The work emphasizes interpretability and reliability through chain-of-thought outputs and iterative refinement, and suggests a path toward more generalizable, context-aware forecasting systems with future work on memory integration and autonomous tool use.

Abstract

Time series forecasting plays a critical role in high-stakes domains such as energy, healthcare, and climate. Although recent advances have improved accuracy, most approaches still treat forecasting as a static one-time mapping task, lacking the interaction, reasoning, and adaptability of human experts. This gap limits their usefulness in complex real-world environments. To address this, we propose AlphaCast, a human wisdom-large language model (LLM) intelligence co-reasoning framework that redefines forecasting as an interactive process. The key idea is to enable step-by-step collaboration between human wisdom and LLM intelligence to jointly prepare, generate, and verify forecasts. The framework consists of two stages: (1) automated prediction preparation, where AlphaCast builds a multi-source cognitive foundation comprising a feature set that captures key statistics and time patterns, a domain knowledge base distilled from corpora and historical series, a contextual repository that stores rich information for each time window, and a case base that retrieves optimal strategies via pattern clustering and matching; and (2) generative reasoning and reflective optimization, where AlphaCast integrates statistical temporal features, prior knowledge, contextual information, and forecasting strategies, triggering a meta-reasoning loop for continuous self-correction and strategy refinement. Extensive experiments on short- and long-term datasets show that AlphaCast consistently outperforms state-of-the-art baselines in predictive accuracy. Code is available at this repository: https://github.com/SkyeGT/AlphaCast_Official .

Figures (7)

• Figure 1: Illustration of the main idea of the newly proposed AlphaCast framework.
• Figure 2: Overall architecture of AlphaCast,which comprises three agents that work in a coordinated manner to accomplish the time-series prediction task.
• Figure 3: Investigate whether the absence of reflection impacts forecasting performance.
• Figure 4: Effect of Exogenous Variables, Timestamps, and Attributes on Model Performance.
• Figure 5: Compare the prediction performance of different reflection mechanisms across different datasets.