Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Multi-relational Graph Diffusion Neural Network with Parallel Retention for Stock Trends Classification

Zinuo You, Pengju Zhang, Jin Zheng, John Cartlidge

TL;DR

This work addresses multi-stock trend classification by modeling the stock market as a dynamic, multi-relational graph. It introduces MGDPR, which generates time-varying, directed edges using edge metrics based on information entropy $H(\cdot)$ and signal energy $E(\cdot)$, refines them through a learnable diffusion process to emphasize task-relevant connections, and employs a parallel retention-based decoupled representation to preserve intra-stock temporal patterns and long-range dependencies. Empirical results on NASDAQ, NYSE, and SSE across seven years demonstrate state-of-the-art performance on next-day trend forecasting, with code released for reproducibility. The contributions include entropy-energy based dynamic graph construction, adaptive graph diffusion, and parallel retention for robust graph representations, advancing practical forecasting in financial markets.

Abstract

Stock trend classification remains a fundamental yet challenging task, owing to the intricate time-evolving dynamics between and within stocks. To tackle these two challenges, we propose a graph-based representation learning approach aimed at predicting the future movements of multiple stocks. Initially, we model the complex time-varying relationships between stocks by generating dynamic multi-relational stock graphs. This is achieved through a novel edge generation algorithm that leverages information entropy and signal energy to quantify the intensity and directionality of inter-stock relations on each trading day. Then, we further refine these initial graphs through a stochastic multi-relational diffusion process, adaptively learning task-optimal edges. Subsequently, we implement a decoupled representation learning scheme with parallel retention to obtain the final graph representation. This strategy better captures the unique temporal features within individual stocks while also capturing the overall structure of the stock graph. Comprehensive experiments conducted on real-world datasets from two US markets (NASDAQ and NYSE) and one Chinese market (Shanghai Stock Exchange: SSE) validate the effectiveness of our method. Our approach consistently outperforms state-of-the-art baselines in forecasting next trading day stock trends across three test periods spanning seven years. Datasets and code have been released (https://github.com/pixelhero98/MGDPR).

Multi-relational Graph Diffusion Neural Network with Parallel Retention for Stock Trends Classification

TL;DR

This work addresses multi-stock trend classification by modeling the stock market as a dynamic, multi-relational graph. It introduces MGDPR, which generates time-varying, directed edges using edge metrics based on information entropy and signal energy , refines them through a learnable diffusion process to emphasize task-relevant connections, and employs a parallel retention-based decoupled representation to preserve intra-stock temporal patterns and long-range dependencies. Empirical results on NASDAQ, NYSE, and SSE across seven years demonstrate state-of-the-art performance on next-day trend forecasting, with code released for reproducibility. The contributions include entropy-energy based dynamic graph construction, adaptive graph diffusion, and parallel retention for robust graph representations, advancing practical forecasting in financial markets.

Abstract

Stock trend classification remains a fundamental yet challenging task, owing to the intricate time-evolving dynamics between and within stocks. To tackle these two challenges, we propose a graph-based representation learning approach aimed at predicting the future movements of multiple stocks. Initially, we model the complex time-varying relationships between stocks by generating dynamic multi-relational stock graphs. This is achieved through a novel edge generation algorithm that leverages information entropy and signal energy to quantify the intensity and directionality of inter-stock relations on each trading day. Then, we further refine these initial graphs through a stochastic multi-relational diffusion process, adaptively learning task-optimal edges. Subsequently, we implement a decoupled representation learning scheme with parallel retention to obtain the final graph representation. This strategy better captures the unique temporal features within individual stocks while also capturing the overall structure of the stock graph. Comprehensive experiments conducted on real-world datasets from two US markets (NASDAQ and NYSE) and one Chinese market (Shanghai Stock Exchange: SSE) validate the effectiveness of our method. Our approach consistently outperforms state-of-the-art baselines in forecasting next trading day stock trends across three test periods spanning seven years. Datasets and code have been released (https://github.com/pixelhero98/MGDPR).
Paper Structure (17 sections, 6 equations, 2 figures, 2 tables)

This paper contains 17 sections, 6 equations, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Attention-based Models for Stock Prediction
  4. GNN-based Models for Complex Networks
  5. Preliminary
  6. Notations
  7. Problem Formulation
  8. Methodology
  9. Dynamic Multi-relational Stock Graph Generation
  10. Multi-relational Graph Diffusion
  11. Graph Representation Learning with Parallel Retention
  12. Objective Function
  13. Experiments
  14. Experiment Setup
  15. Experimental Result and Analysis
  16. ...and 2 more sections

Figures (2)

  • Figure 1: Schematic of graph representation learning scheme.
  • Figure 2: Ablation study results. Only the full MGDPR (P1+P2+P3) has higher accuracy than the best baseline in each market.