Paper: arXiv 2402.06633

Abstract

The stock market is a crucial component of the financial system, but predicting the movement of stock prices is challenging due to the dynamic and intricate relations arising from various aspects such as economic indicators, financial reports, global news, and investor sentiment. Traditional sequential methods and graph-based models have been applied in stock movement prediction, but they have limitations in capturing the multifaceted and temporal influences in stock price movements. To address these challenges, the Multi-relational Dynamic Graph Neural Network (MDGNN) framework is proposed, which utilizes a discrete dynamic graph to comprehensively capture multifaceted relations among stocks and their evolution over time. The representation generated from the graph offers a complete perspective on the interrelationships among stocks and associated entities. Additionally, the power of the Transformer structure is leveraged to encode the temporal evolution of multiplex relations, providing a dynamic and effective approach to predicting stock investment. Further, our proposed MDGNN framework achieves the best performance in public datasets compared with state-of-the-art (SOTA) stock investment methods.

Complexity vs Empirical Score

  • Math Complexity: 7.5/10
  • Empirical Rigor: 6.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced machine learning techniques involving dynamic graph neural networks and Transformers, requiring significant mathematical formalism, but is heavily reliant on public datasets with reported benchmark comparisons indicating substantial implementation and backtesting efforts.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Predict Stock Movements<br>with Dynamic Relations"] --> B["Data & Inputs"]
  B --> C["Core Methodology: MDGNN"]
  
  subgraph B ["Data / Inputs"]
      B1["Stock Price Data"]
      B2["Economic Indicators"]
      B3["News & Sentiment"]
  end

  subgraph C ["MDGNN Framework"]
      C1["Discrete Dynamic Graph Construction<br> Captures multifaceted relations"]
      C2["Multi-relational GNN<br> Learns node embeddings"]
      C3["Transformer Encoder<br> Models temporal evolution"]
  end

  C --> D["Computational Process"]
  D --> E["Key Outcomes"]

  subgraph D ["Process Flow"]
      direction LR
      D1["Graph Generation"]
      D2["Spatial Aggregation"]
      D3["Temporal Encoding"]
  end

  subgraph E ["Findings"]
      E1["State-of-the-Art Performance"]
      E2["Comprehensive Relation Capture"]
      E3["Dynamic Investment Strategy"]
  end

Cited References (214)

  1. Blackboard Systems (1986)
  2. Communication, Simulation, and Intelligent Agents: Implications of Personal Intelligent Machines for Medical Education (1983)
  3. Classification Problem Solving (1984)
  4. New Ways to Make Microcircuits Smaller (1980) — link
  5. New Ways to Make Microcircuits Smaller—Duplicate Entry (1980)
  6. Strategic explanations for a diagnostic consultation system (1984) — link
  7. Strategic Explanations in Consultation—Duplicate (1983)
  8. Poligon: A System for Parallel Problem Solving (1986)
and 206 more
  1. Transfer of Rule-Based Expertise through a Tutorial Dialogue (1979)
  2. The Engineering of Qualitative Models (2021)
  3. Attention Is All You Need (2017) — link
  4. Pluto: The ‘Other’ Red Planet (2015)
  5. Wide & Deep Learning for Recommender Systems (2016) — link
  6. Attention is All you Need (2017)
  7. Long Short-Term Memory (1997) — link
  8. Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling (2014) — link
  9. Deep & Cross Network for Ad Click Predictions (2017) — link
  10. Ad Click Prediction: A View from the Trenches (2013) — link
  11. Residual Attention Network for Image Classification (2017) — link
  12. BERT : Pre-training of Deep Bidirectional Transformers for Language Understanding (2019) — link
  13. Learning to Rank Using Gradient Descent (2005) — link
  14. Learning to Rank with Nonsmooth Cost Functions (2007)
  15. Learning to Rank: From Pairwise Approach to Listwise Approach (2007) — link
  16. Learning a Deep Listwise Context Model for Ranking Refinement (2018) — link
  17. Personalized Re-Ranking for Recommendation (2019) — link
  18. Seq2Slate: Re-ranking and Slate Optimization with RNNs (2019) — link
  19. Globally Optimized Mutual Influence Aware Ranking in E-Commerce Search (2018)
  20. SetRank: Learning a Permutation-Invariant Ranking Model for Information Retrieval (2020) — link
  21. Recurrent Neural Networks: Design and Applications (1999)
  22. On Layer Normalization in the Transformer Architecture (2020) — link
  23. Yahoo! Learning to Rank Challenge Overview (2010)
  24. Introducing LETOR 4.0 Datasets (2013)
  25. Layer Normalization (2016) — link
  26. From RankNet to LambdaRank to LambdaMART: An Overview (2010) — link
  27. Multi-Scale Self-Attention for Text Classification (2020) — link
  28. Convolutional Self-Attention Networks (2019) — link
  29. Attention Augmented Convolutional Networks (2019) — link
  30. Massive Exploration of Neural Machine Translation Architectures (2017) — link
  31. AdaRank: A Boosting Algorithm for Information Retrieval (2007) — link
  32. SoftRank: Optimizing Non-Smooth Rank Metrics (2008) — link
  33. IR Evaluation Methods for Retrieving Highly Relevant Documents (2000) — link
  34. A Support Vector Method for Optimizing Average Precision (2007) — link
  35. Deep Residual Learning for Image Recognition (2016) — link
  36. Learning to Rank for Information Retrieval (2009) — link
  37. Two-Stage Learning to Rank for Information Retrieval (2013) — link
  38. Learning to Rank for Recommender Systems (2013) — link
  39. Learning to Rank by Optimizing NDCG Measure (2009)
  40. Learning Groupwise Multivariate Scoring Functions Using Deep Neural Networks (2019) — link
  41. Whole Page Optimization with Global Constraints (2019) — link
  42. Deep Reinforcement Learning for Page-Wise Recommendations (2018) — link
  43. Training Linear SVMs in Linear Time (2006) — link
  44. On Position Embeddings in Bert (2021)
  45. What Do Position Embeddings Learn? An Empirical Study of Pre-Trained Language Model Positional Encoding (2020) — link
  46. Grid-Based Evaluation Metrics for Web Image Search (2019) — link
  47. Ads Allocation in Feed via Constrained Optimization (2020) — link
  48. Debiasing Grid-Based Product Search in E-Commerce (2020) — link
  49. TF-Ranking: Scalable TensorFlow Library for Learning-to-Rank (2019) — link
  50. TensorFlow: A System for Large-Scale Machine Learning (2016)
  51. Language Models are Few-Shot Learners (2020)
  52. Improving Language Understanding by Generative Pre-Training (2018)
  53. Practical Diversified Recommendations on YouTube with Determinantal Point Processes (2018) — link
  54. Bias and Debias in Recommender System: A Survey and Future Directions (2020) — link
  55. SetRank: Learning a Permutation-Invariant Ranking Model for Information Retrieval (2020) — link
  56. Are Neural Rankers still Outperformed by Gradient Boosted Decision Trees? (2021)
  57. The LambdaLoss Framework for Ranking Metric Optimization (2018) — link
  58. Are Neural Rankers still Outperformed by Gradient Boosted Decision Trees? (2021)
  59. Feature Transformation for Neural Ranking Models (2020) — link
  60. Multi-Interactive Attention Network for Fine-Grained Feature Learning in CTR Prediction (2021) — link
  61. SIFN: A Sentiment-Aware Interactive Fusion Network for Review-Based Item Recommendation (2021) — link
  62. Graph Contextualized Self-Attention Network for Session-based Recommendation (2019) — link
  63. Session-Based Recommendation with Graph Neural Networks (2019) — link
  64. Rethinking the Item Order in Session-Based Recommendation with Graph Neural Networks (2019) — link
  65. Memory Augmented Graph Neural Networks for Sequential Recommendation (2020) — link
  66. Global Context Enhanced Graph Neural Networks for Session-Based Recommendation (2020) — link
  67. Deep Interest Network for Click-Through Rate Prediction (2018) — link
  68. Deep Interest Evolution Network for Click-Through Rate Prediction (2019) — link
  69. Practice on Long Sequential User Behavior Modeling for Click-Through Rate Prediction (2019) — link
  70. Search-Based User Interest Modeling with Lifelong Sequential Behavior Data for Click-Through Rate Prediction (2020) — link
  71. An introduction to ROC analysis (2006)
  72. Semi-Supervised Classification with Graph Convolutional Networks (2017) — link
  73. Graph Attention Networks (2018) — link
  74. LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation (2020) — link
  75. GeniePath: Graph Neural Networks with Adaptive Receptive Paths (2019) — link
  76. Graph Neural Networks in Recommender Systems: A Survey (2021) — link
  77. Field-aware factorization machines for CTR prediction (2016)
  78. Field-weighted factorization machines for click-through rate prediction in display advertising (2018)
  79. DeepFM: a factorization-machine based neural network for CTR prediction (2017) — link
  80. xdeepfm: Combining explicit and implicit feature interactions for recommender systems (2018)
  81. Sparse factorization machines for click-through rate prediction (2016)
  82. Factorization Machines (2010) — link
  83. Learning from history and present: Next-item recommendation via discriminatively exploiting user behaviors (2018)
  84. Vista: A visually, socially, and temporally-aware model for artistic recommendation (2016)
  85. Translation-based recommendation (2017)
  86. Fusing similarity models with markov chains for sparse sequential recommendation (2016)
  87. Factorizing personalized markov chains for next-basket recommendation (2010)
  88. Personalized top-n sequential recommendation via convolutional sequence embedding (2018)
  89. Learning hierarchical representation model for nextbasket recommendation (2015)
  90. Recurrent recommender networks (2017)
  91. Recurrent neural networks with top-k gains for session-based recommendations (2018)
  92. Gradient flow in recurrent nets: the difficulty of learning long-term dependencies (2001)
  93. Deep Neural Networks for YouTube Recommendations (2016) — link
  94. CAN: Revisiting Feature Co-Action for Click-Through Rate Prediction (2020) — link
  95. Session-based Recommendations with Recurrent Neural Networks (2016) — link
  96. BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer (2019) — link
  97. Behavior Sequence Transformer for E-Commerce Recommendation in Alibaba (2019) — link
  98. Multi-Interest Network with Dynamic Routing for Recommendation at Tmall (2019) — link
  99. Deep Session Interest Network for Click-through Rate Prediction (2019)
  100. Deep Learning using Rectified Linear Units (ReLU) (2019) — link
  101. Self-Attentive Sequential Recommendation (2018) — link
  102. Personalized Top-N Sequential Recommendation via Convolutional Sequence Embedding (2018) — link
  103. CSAN: Contextual Self-Attention Network for User Sequential Recommendation (2018) — link
  104. Dropout: A Simple Way to Prevent Neural Networks from Overfitting (2014)
  105. Latent Cross: Making Use of Context in Recurrent Recommender Systems (2018) — link
  106. Comparison of Implicit and Explicit Feedback from an Online Music Recommendation Service (2010) — link
  107. Collaborative Filtering for Implicit Feedback Datasets (2008) — link
  108. Deep Feedback Network for Recommendation (2021)
  109. Modeling User Exposure in Recommendation (2015)
  110. Fast Matrix Factorization for Online Recommendation with Implicit Feedback (2016) — link
  111. Explicit or Implicit Feedback? Engagement or Satisfaction? A Field Experiment on Machine-Learning-Based Recommender Systems (2018) — link
  112. FeedRec: News Feed Recommendation with Various User Feedbacks (2022) — link
  113. Denoising User-Aware Memory Network for Recommendation (2021) — link
  114. Curriculum Disentangled Recommendation with Noisy Multi-feedback (2021) — link
  115. M ix H op: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing (2019) — link
  116. Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs (2020) — link
  117. Universal Graph Convolutional Networks (2021) — link
  118. Geom-GCN: Geometric Graph Convolutional Networks (2020) — link
  119. Non-Local Graph Neural Networks (5555) — link
  120. Graph Pointer Neural Networks (2022) — link
  121. Powerful Graph Convolutional Networks with Adaptive Propagation Mechanism for Homophily and Heterophily (2022) — link
  122. Beyond Low-frequency Information in Graph Convolutional Networks (2021) — link
  123. Simple and Deep Graph Convolutional Networks (2020) — link
  124. Adaptive Universal Generalized PageRank Graph Neural Network (2021) — link
  125. Breaking the Limit of Graph Neural Networks by Improving the Assortativity of Graphs with Local Mixing Patterns (2021) — link
  126. Diverse Message Passing for Attribute with Heterophily (2021) — link
  127. Ups and Downs: Modeling the Visual Evolution of Fashion Trends with One-Class Collaborative Filtering (2016) — link
  128. The MovieLens Datasets: History and Context (2015) — link
  129. Explainable Reasoning over Knowledge Graphs for Recommendation (2019) — link
  130. Knowledge Graph Convolutional Networks for Recommender Systems (2019) — link
  131. Simplifying Graph Convolutional Networks (2019) — link
  132. Label Efficient Semi-Supervised Learning via Graph Filtering (2019)
  133. AM-GCN: Adaptive Multi-Channel Graph Convolutional Networks (2020) — link
  134. Measuring and Relieving the Over-Smoothing Problem for Graph Neural Networks from the Topological View (2020) — link
  135. Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning (2018)
  136. KGAT: Knowledge Graph Attention Network for Recommendation (2019) — link
  137. A survey of heterogeneous information network analysis (2017) — link
  138. Multi-View Denoising Graph Auto-Encoders on Heterogeneous Information Networks for Cold-Start Recommendation (2021) — link
  139. Relevance Search in Heterogeneous Networks (2012) — link
  140. PathSim: Meta Path-Based Top-K Similarity Search in Heterogeneous Information Networks (2011) — link
  141. PathSelClus: Integrating Meta-Path Selection with User-Guided Object Clustering in Heterogeneous Information Networks (2013) — link
  142. Meta Path-Based Collective Classification in Heterogeneous Information Networks (2012) — link
  143. Heterogeneous Graph Transformer (2020) — link
  144. Heterogeneous Graph Attention Network (2019) — link
  145. Neural Multi-task Recommendation from Multi-behavior Data (2019) — link
  146. Multiplex Behavioral Relation Learning for Recommendation via Memory Augmented Transformer Network (2020) — link
  147. Multi-Behavior Recommendation with Graph Convolutional Networks (2020) — link
  148. Efficient Heterogeneous Collaborative Filtering without Negative Sampling for Recommendation (2020) — link
  149. Graph Meta Network for Multi-Behavior Recommendation (2021) — link
  150. Graph Heterogeneous Multi-Relational Recommendation (2021) — link
  151. Knowledge-Enhanced Hierarchical Graph Transformer Network for Multi-Behavior Recommendation (2021) — link
  152. Contrastive Meta Learning with Behavior Multiplicity for Recommendation (2022) — link
  153. Multi-Behavior Hypergraph-Enhanced Transformer for Sequential Recommendation (2022) — link
  154. Self-supervised Graph Neural Networks for Multi-behavior Recommendation (2022) — link
  155. dyngraph2vec: Capturing network dynamics using dynamic graph representation learning (2020) — link
  156. EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphs (2020) — link
  157. Inductive representation learning on temporal graphs (2020) — link
  158. Network Embedding and Change Modeling in Dynamic Heterogeneous Networks (2019) — link
  159. Predicting Dynamic Embedding Trajectory in Temporal Interaction Networks (2019) — link
  160. DepthLGP: Learning Embeddings of Out-of-Sample Nodes in Dynamic Networks (2018) — link
  161. Dynamic Network Embedding : An Extended Approach for Skip-gram based Network Embedding (2018) — link
  162. DynGEM: Deep Embedding Method for Dynamic Graphs (2018) — link
  163. NetWalk: A Flexible Deep Embedding Approach for Anomaly Detection in Dynamic Networks (2018) — link
  164. DeeperGCN: All You Need to Train Deeper GCNs (2020) — link
  165. How Powerful are Graph Neural Networks? (2019) — link
  166. Weisfeiler and Leman Go Neural: Higher-Order Graph Neural Networks (2019) — link
  167. On the Bottleneck of Graph Neural Networks and its Practical Implications (2021) — link
  168. Do Transformers Really Perform Badly for Graph Representation? (2021) — link
  169. Structure-Aware Transformer for Graph Representation Learning (2022)
  170. GraphiT: Encoding Graph Structure in Transformers (2021) — link
  171. Graph-Based Learning for Stock Movement Prediction with Textual and Relational Data (2021)
  172. HIST: A Graph-based Framework for Stock Trend Forecasting via Mining Concept-Oriented Shared Information (2021) — link
  173. Hierarchical Adaptive Temporal-Relational Modeling for Stock Trend Prediction (2021) — link
  174. Stock Selection via Spatiotemporal Hypergraph Attention Network: A Learning to Rank Approach (2021) — link
  175. Adaptive Long-Short Pattern Transformer for Stock Investment Selection (2022) — link
  176. DySAT: Deep Neural Representation Learning on Dynamic Graphs via Self-Attention Networks (2020)
  177. ARIMA Models (1990) — link
  178. Temporal Relational Ranking for Stock Prediction (2018) — link
  179. PyTorch: An Imperative Style, High-Performance Deep Learning Library (2019) — link
  180. Modeling the Stock Relation with Graph Network for Overnight Stock Movement Prediction (2020) — link
  181. Temporal Relational Ranking for Stock Prediction (2019) — link
  182. Individualized Indicator for All: Stock-Wise Technical Indicator Optimization with Stock Embedding (2019) — link
  183. Heterogeneous Temporal Graph Neural Network (2021) — link
  184. RoFormer: Enhanced Transformer with Rotary Position Embedding (2022) — link
  185. Modeling Relational Data with Graph Convolutional Networks (2017) — link
  186. Hierarchical Multi-Scale Gaussian Transformer for Stock Movement Prediction (2021)
  187. Learning the Evolutionary and Multi-Scale Graph Structure for Multivariate Time Series Forecasting (2022) — link
  188. Incorporating Expert-Based Investment Opinion Signals in Stock Prediction: A Deep Learning Framework (2020) — link
  189. Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation (2022) — link
  190. Enhancing Stock Movement Prediction with Adversarial Training (2019) — link
  191. Learning Multiple Stock Trading Patterns with Temporal Routing Adaptor and Optimal Transport (2021) — link
  192. Stock Price Prediction via Discovering Multi-Frequency Trading Patterns (2017) — link
  193. Machine learning in asset pricing (2021)
  194. Empirical asset pricing via machine learning (2020)
  195. Stock market forecasting using LASSO linear regression model (2015)
  196. A machine learning trading system for the stock market based on N-period Min-Max labeling using XGBoost (2023)
  197. Stock market’s price movement prediction with LSTM neural networks (2017)
  198. Semi-supervised classification with graph convolutional networks (2016) — link
  199. Heterogeneous graph neural network (2019)
  200. Heterogeneous graph transformer (2020)
  201. Long-term spatio-temporal forecasting via dynamic multiple-graph attention (2022) — link
  202. Heterogeneous temporal graph neural network (2022)
  203. Temporal relational ranking for stock prediction (2019)
  204. Graph-based learning for stock movement prediction with textual and relational data (2021) — link
  205. Hierarchical Adaptive Temporal-Relational Modeling for Stock Trend Prediction. (2021)
  206. Hist: A graph-based framework for stock trend forecasting via mining concept-oriented shared information (2021) — link

0 of 214 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.