SPH-Net: A Co-Attention Hybrid Model for Accurate Stock Price Prediction

2025-09-18

AI Summary

Stock price forecasting remains highly challenging due to the high volatility, non-stationarity, and strong nonlinearity inherent in financial time series. To address these challenges, we propose a novel deep temporal modeling framework that integrates Vision Transformers (ViT) with a custom collaborative attention mechanism. Our approach reshapes one-dimensional stock price sequences into two-dimensional pseudo-images for ViT input, enabling simultaneous global dependency modeling and local pattern extraction via cross-scale attention, while enhancing robustness to noise. The model is trained end-to-end on six daily features—open, high, low, close, adjusted close, and volume—for eight stocks. Extensive experiments across multiple empirical datasets demonstrate consistent superiority over baseline models—including LSTM, TCN, Informer, and standard ViT—achieving state-of-the-art performance in MAE, RMSE, and Direction Accuracy. These results validate both the effectiveness and generalizability of our framework for financial time series forecasting.

Technology Category

• Computer Vision: Large Vision Models
• Machine Learning: Time-Series/Data Streams
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• 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
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Prediction of stock price movements presents a formidable challenge in financial analytics due to the inherent volatility, non-stationarity, and nonlinear characteristics of market data. This paper introduces SPH-Net (Stock Price Prediction Hybrid Neural Network), an innovative deep learning framework designed to enhance the accuracy of time series forecasting in financial markets. The proposed architecture employs a novel co-attention mechanism that initially processes temporal patterns through a Vision Transformer, followed by refined feature extraction via an attention mechanism, thereby capturing both global and local dependencies in market data. To rigorously evaluate the model's performance, we conduct comprehensive experiments on eight diverse stock datasets: AMD, Ebay, Facebook, FirstService Corp, Tesla, Google, Mondi ADR, and Matador Resources. Each dataset is standardized using six fundamental market indicators: Open, High, Low, Close, Adjusted Close, and Volume, representing a complete set of features for comprehensive market analysis. Experimental results demonstrate that SPH-Net consistently outperforms existing stock prediction models across all evaluation metrics. The model's superior performance stems from its ability to effectively capture complex temporal patterns while maintaining robustness against market noise. By significantly improving prediction accuracy in financial time series analysis, SPH-Net provides valuable decision-support capabilities for investors and financial analysts, potentially enabling more informed investment strategies and risk assessment in volatile market conditions.

Problem

• Predicting stock price movements using deep learning
• Capturing global and local dependencies in market data
• Improving accuracy in financial time series analysis

Innovation

• Co-attention hybrid model combining Vision Transformer
• Refined feature extraction via attention mechanism
• Captures global and local market dependencies