Cross-Modal Temporal Fusion for Financial Market Forecasting

2025-04-18

AI Summary

To address the limited prediction accuracy in multi-source heterogeneous financial forecasting—caused by insufficient intra-modal temporal modeling and inter-modal interaction under conventional modality-isolated approaches—this paper proposes a Dynamic Cross-Modal Temporal Fusion (DCMTF) framework. DCMTF introduces a novel cross-modal attention weighting mechanism and an interpretable tensor decomposition module to jointly model intra-modal temporal dependencies and dynamically evolving inter-modal interactions, enabling contribution-aware attribution and feature disentanglement. The framework integrates self-supervised pretraining, multi-head cross-modal attention, temporal tensor decomposition, and automated hyperparameter optimization into an end-to-end, production-ready training pipeline. Evaluated on real-world market data, DCMTF achieves an average 4.2% improvement in stock price movement prediction accuracy over state-of-the-art baselines, while demonstrating strong scalability and deployment feasibility.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Accurate financial market forecasting requires diverse data sources, including historical price trends, macroeconomic indicators, and financial news, each contributing unique predictive signals. However, existing methods often process these modalities independently or fail to effectively model their interactions. In this paper, we introduce Cross-Modal Temporal Fusion (CMTF), a novel transformer-based framework that integrates heterogeneous financial data to improve predictive accuracy. Our approach employs attention mechanisms to dynamically weight the contribution of different modalities, along with a specialized tensor interpretation module for feature extraction. To facilitate rapid model iteration in industry applications, we incorporate a mature auto-training scheme that streamlines optimization. When applied to real-world financial datasets, CMTF demonstrates improvements over baseline models in forecasting stock price movements and provides a scalable and effective solution for cross-modal integration in financial market prediction.

Problem

• Integrates diverse financial data for market forecasting
• Models interactions between different data modalities effectively
• Improves stock price prediction accuracy with cross-modal fusion

Innovation

• Transformer-based framework integrates diverse financial data
• Attention mechanisms dynamically weight modality contributions
• Auto-training scheme streamlines model optimization process