Multi-period Learning for Financial Time Series Forecasting

2025-07-20

Knowledge Discovery and Data Mining

AI Summary

Financial time-series forecasting faces challenges in effectively integrating multi-horizon information—such as short-term sentiment, medium-to-long-term policy shifts, and macro market trends. To address this, we propose the Multi-Horizon Learning Framework (MLF), a novel architecture featuring inter-horizon redundancy filtering, learnable weighted fusion, and adaptive segmentation. MLF further incorporates Patch Squeeze—a lightweight compression technique—and an enhanced self-attention module to enable efficient modeling of variable-length multi-horizon inputs. Compared to state-of-the-art methods, MLF achieves superior prediction accuracy, reduces reliance on fixed input lengths, and lowers training configuration overhead. Extensive experiments across multiple financial datasets demonstrate its effectiveness and strong generalization capability. The source code and datasets are publicly available.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Computer Vision: Multi-modal Vision
• Multiagent Systems: Multiagent Learning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Time series forecasting is important in finance domain. Financial time series (TS) patterns are influenced by both short-term public opinions and medium-/long-term policy and market trends. Hence, processing multi-period inputs becomes crucial for accurate financial time series forecasting (TSF). However, current TSF models either use only single-period input, or lack customized designs for addressing multi-period characteristics. In this paper, we propose a Multi-period Learning Framework (MLF) to enhance financial TSF performance. MLF considers both TSF's accuracy and efficiency requirements. Specifically, we design three new modules to better integrate the multi-period inputs for improving accuracy: (i) Inter-period Redundancy Filtering (IRF), that removes the information redundancy between periods for accurate self-attention modeling, (ii) Learnable Weighted-average Integration (LWI), that effectively integrates multi-period forecasts, (iii) Multi-period self-Adaptive Patching (MAP), that mitigates the bias towards certain periods by setting the same number of patches across all periods. Furthermore, we propose a Patch Squeeze module to reduce the number of patches in self-attention modeling for maximized efficiency. MLF incorporates multiple inputs with varying lengths (periods) to achieve better accuracy and reduces the costs of selecting input lengths during training. The codes and datasets are available at https://github.com/Meteor-Stars/MLF.

Problem

• Forecasting financial time series using multi-period inputs
• Addressing redundancy and bias in multi-period data integration
• Balancing accuracy and efficiency in time series modeling

Innovation

• Inter-period Redundancy Filtering removes redundancy between periods
• Learnable Weighted-average Integration combines multi-period forecasts
• Multi-period self-Adaptive Patching mitigates period bias uniformly