Paper: arXiv 2402.18959

Abstract

The stock market plays a pivotal role in economic development, yet its intricate volatility poses challenges for investors. Consequently, research and accurate predictions of stock price movements are crucial for mitigating risks. Traditional time series models fall short in capturing nonlinearity, leading to unsatisfactory stock predictions. This limitation has spurred the widespread adoption of neural networks for stock prediction, owing to their robust nonlinear generalization capabilities. Recently, Mamba, a structured state space sequence model with a selection mechanism and scan module (S6), has emerged as a powerful tool in sequence modeling tasks. Leveraging this framework, this paper proposes a novel Mamba-based model for stock price prediction, named MambaStock. The proposed MambaStock model effectively mines historical stock market data to predict future stock prices without handcrafted features or extensive preprocessing procedures. Empirical studies on several stocks indicate that the MambaStock model outperforms previous methods, delivering highly accurate predictions. This enhanced accuracy can assist investors and institutions in making informed decisions, aiming to maximize returns while minimizing risks. This work underscores the value of Mamba in time-series forecasting. Source code is available at https://github.com/zshicode/MambaStock.

Complexity vs Empirical Score

  • Math Complexity: 7.5/10
  • Empirical Rigor: 4.0/10
  • Quadrant: Lab Rats — theoretically deep, empirically untested

Why this score: The paper introduces advanced state-space models (S4/Mamba) with derived equations and discrete-time transitions, requiring moderate mathematical sophistication. However, empirical rigor is limited by vague performance metrics, small dataset specifics, and lack of detailed backtesting methodology or risk metrics, despite providing code and data sources.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Stock Prediction"] --> B["Input Data:<br>Historical Stock Prices"]
  B --> C["Method: MambaStock<br>Structured State Space Model S6"]
  C --> D{"Computation:<br>Sequence Modeling & Forecasting"}
  D --> E["Output:<br>Future Stock Prices"]
  E --> F["Outcome:<br>SOTA Accuracy & Low Risk"]

Cited References (183)

  1. Semi-supervised drug-protein interaction prediction from heterogeneous biological spaces (2010)
  2. An Interpretable Double-Scale Attention Model for Enzyme Protein Class Prediction Based on Transformer Encoders and Multi-Scale Convolutions (2022)
  3. A Singular Value Thresholding Algorithm for Matrix Completion (2010)
  4. On the $O(1/n)$ Convergence Rate of the Douglas-Rachford Alternating Direction Method (2012)
  5. The Direct Extension of ADMM for Multi-Block Convex Minimization Problems is Not Necessarily Convergent (2016)
  6. A method of solving a convex programming problem with convergence rate O(1/k2) (1983)
  7. An Accelerated Proximal Gradient Algorithm for Nuclear Norm Regularized Least Squares Problems (2010)
  8. Learning with Hypergraphs: Clustering, Classification, and Embedding (2006)
and 175 more
  1. Sparse Subspace Clustering (2009)
  2. Solutions of ill-posed problems (1977)
  3. A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems (2009)
  4. Robust Principal Component Analysis? (2011)
  5. An Iterative Thresholding Algorithm for Linear Inverse Problems with a Sparsity Constrains (2004)
  6. Global Convergence of Unmodified 3-Block ADMM for a Class of Convex Minimization Problems (2018)
  7. Stochastic Analysis on Manifolds (2002)
  8. A Concise and Provably Informative Multi-Scale Signature Based on Heat Diffusion (2009)
  9. Differential equation and probability inspired graph neural networks for latent variable learning (2022) — link
  10. Diffusion kernels on graphs and other discrete input spaces (2002)
  11. Robust Recovery of Subspace Structures by Low-Rank Representation (2013)
  12. Linearized Alternating Direction Method with Adaptive Penalty and Warm Starts for Fast Solving Transform Invariant Low-Rank Textures (2013)
  13. PTE: Predictive Text Embedding through Large-scale Heterogeneous Text Networks (2015)
  14. Graph Convolutional Networks for Text Classification (2019)
  15. Collective Classification of Network Data (2008)
  16. Continuous Graph Neural Networks (2020)
  17. Connecting the Dots: Identifying Network Structure via Graph Signal Processing (2019)
  18. DropEdge: Towards Deep Graph Convolutional Networks on Node Classification (2020)
  19. Adaptive Graph Encoder for Attributed Graph Embedding (2020)
  20. Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning (2018)
  21. Graph Regularized Nonnegative Matrix Factorization for Data Representation (2011)
  22. Linearized Alternating Direction Method with Adaptive Penalty for Low-Rank Representation (2011)
  23. Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers (2010)
  24. The linear neighborhood propagation method for predicting long non-coding RNA–protein interactions (2018)
  25. Learning with Local and Global Consistency (2004)
  26. GMNN : Graph M arkov Neural Networks (2019)
  27. Variational Inference with Normalizing Flows (2015)
  28. Improved Variational Inference with Inverse Autoregressive Flow (2016)
  29. Haar Graph Pooling (2020)
  30. Spectral Clustering with Graph Neural Networks for Graph Pooling (2020)
  31. Structural Deep Clustering Network (2020)
  32. On the Convergence of Adam and Beyond (2018)
  33. Principal Neighbourhood Aggregation for Graph Nets (2020) — link
  34. Order Matters: Sequence to sequence for sets (2016)
  35. Learning Convolutional Neural Networks for Graphs (2016)
  36. StructPool: Structured Graph Pooling via Conditional Random Fields (2020)
  37. Graph Convolutional Networks with EigenPooling (2019)
  38. A Fair Comparison of Graph Neural Networks for Graph Classification (2020)
  39. Distinguishing Enzyme Structures from Non-enzymes Without Alignments (2003)
  40. Weisfeiler-Lehman Graph Kernels (2011)
  41. Hierarchical Graph Representation Learning withDifferentiable Pooling (2018)
  42. An End-to-End Deep Learning Architecture for Graph Classification (2018)
  43. Self-Attention Graph Pooling (2019)
  44. How Powerful are Graph Neural Networks? (2019)
  45. Graph U-Nets (2019)
  46. Learned Convolutional Sparse Coding (2018)
  47. Spectral Graph Theory (1997)
  48. Graph Convolutional Networks using Heat Kernel for Semi-supervised Learning (2019) — link
  49. Adam: A Method for Stochastic Optimization (2015)
  50. Dropout: A Simple Way to Prevent Neural Networks from Overfitting (2014)
  51. Manifold Regularization: A Geometric Framework for Learning from Labeled and Unlabeled Examples (2006)
  52. Laplacian Eigenmaps and Spectral Techniques for Embedding and Clustering (2001)
  53. DeepWalk: Online Learning of Social Representations (2014)
  54. A tutorial on spectral clustering (2007)
  55. Normalized cuts and image segmentation (2000)
  56. On Spectral Clustering: Analysis and an Algorithm (2001)
  57. The Nature of Statistical Learning Theory (2000)
  58. Numerical Optimization (2006)
  59. Nonlinear Dimensionality Reduction by Locally Linear Embedding (2000)
  60. Distributed Representations of Sentences and Documents (2014)
  61. TPGLDA: Novel prediction of associations between lncRNAs and diseases via lncRNA-disease-gene tripartite graph (2018)
  62. ProtVec: A Continuous Distributed Representation of Biological Sequences (2015)
  63. Semi-Supervised Classification with Graph Convolutional Networks (2017)
  64. Inductive Representation Learning on Large Graphs (2017)
  65. Graph Attention Networks (2018)
  66. Simplifying Graph Convolutional Networks (2019)
  67. Revisiting Graph Neural Networks: All We Have is Low-Pass Filters (2019) — link
  68. Unifying Graph Convolutional Neural Networks and Label Propagation (2020) — link
  69. Graph Wavelet Neural Network (2019)
  70. How Powerful are Graph Neural Networks? (2019)
  71. What Can Neural Networks Reason About? (2020)
  72. A novel computational framework for simultaneous integration of multiple types of genomic data to identify microRNA-gene regulatory modules (2011)
  73. Efficient label propagation for interactive image segmentation (2007)
  74. Discovery of multi-dimensional modules by integrative analysis of cancer genomic data (2012)
  75. On a routing problem (1958)
  76. The Graph Neural Network Model (2009)
  77. Attention-based Graph Neural Network for Semi-supervised Learning (2018) — link
  78. Neural Message Passing for Quantum Chemistry (2017)
  79. A reduction of a graph to a canonical form and an algebraarising during this reduction (1968)
  80. Auto-Encoding Variational Bayes (2013)
  81. Benchmark Data Sets for Graph Kernels (2020) — link
  82. Open Graph Benchmark: Datasets for Machine Learning on Graphs (2020) — link
  83. Variational Graph Auto-Encoders (2016)
  84. Zero-shot Recognition via Semantic Embeddings and Knowledge Graphs (2018)
  85. Learning from labels and unlabeled data with label propagation (2002)
  86. The Augmented Lagrange Multiplier Method for Exact Recovery of Corrupted Low-Rank Matrices (2009)
  87. The PageRank Citation Ranking: Bringing Order to the Web (1999)
  88. Semi-Supervised Learning Using Gaussian Fields and Harmonic Functions (2003)
  89. Label Efficient Semi-Supervised Learning via Graph Filtering (2019)
  90. Zero-Shot Fine-Grained Classification by Deep Feature Learning with Semantics (2017) — link
  91. Robust Subspace Segmentation by Low-Rank Representation (2010)
  92. Robust Recovery of Subspace Structures by Low-Rank Representation (2010) — link
  93. Latent Low-Rank Representation for subspace segmentation and feature extraction (2011)
  94. Learning to Propagate Labels: Transductive Propagation Network for Few-shot Learning (2019)
  95. Label Information Guided Graph Construction for Semi-Supervised Learning (2017)
  96. Algorithms for non-negative matrix factorization (2000)
  97. Predicting CircRNA-Disease Associations Through Linear Neighborhood Label Propagation Method (2019)
  98. Non-negative Laplacian Embedding (2009)
  99. Linear Neighborhood Propagation and Its Applications (2009)
  100. Label Propagation through Linear Neighborhoods (2006)
  101. A Proposal on Machine Learning via Dynamical Systems (2017)
  102. Beyond Finite Layer Neural Networks: Bridging Deep Architectures and Numerical Differential Equations (2018)
  103. An Optimal Control Approach to Deep Learning and Applications to Discrete-Weight Neural Networks (2018)
  104. Maximum Principle Based Algorithms for Deep Learning (2017)
  105. Neural Ordinary Differential Equations (2018)
  106. Global Linear Neighborhoods for Efficient Label Propagation (2012)
  107. Sequence-derived linear neighborhood propagation method for predicting lncRNA-miRNA interactions (2018)
  108. Incorporating Transformer and LSTM to Kalman Filter with EM algorithm for state estimation (2021) — link
  109. Bidirectional LSTM-CRF Attention-based Model for Chinese Word Segmentation (2021) — link
  110. Time series forecasting using a hybrid ARIMA and neural network model (2003)
  111. Predicting miRNA-Disease Association Based on Neural Inductive Matrix Completion with Graph Autoencoders and Self-Attention Mechanism (2022)
  112. Transformer-XL: attentive language models beyond a fixed-length context (2019) — link
  113. Self-attention with relative position representations (2018)
  114. How much attention do you need? a granular analysis of neural machine translation architectures (2018)
  115. MX-LSTM: mixing tracklets and vislets to jointly forecast trajectories and head poses (2018)
  116. Hybrid neural networks for learning the trend in time series (2017)
  117. ATTCry: Attention-based neural network model for protein crystallization prediction (2021)
  118. TLCrys: Transfer Learning Based Method for Protein Crystallization Prediction (2022)
  119. Predicting lncRNA–protein interactions based on graph autoencoders and collaborative training (2021)
  120. Backpropagation applied to handwritten zip code recognition (1989)
  121. Principal component analysis (1987)
  122. Generative adversarial networks (2014) — link
  123. Deep residual learning for image recognition (2016)
  124. Neural machine translation by jointly learning to align and translate (2015)
  125. Optimal state estimation: Kalman, H Infinity, and nonlinear approaches (2006)
  126. Time series analysis: forecasting and control (1970)
  127. State estimation for robotics (2017)
  128. Gaussian mixture sigma-point particle filters for sequential probabilistic inference in dynamic state-space models (2003)
  129. A representation learning model based on variational inference and graph autoencoder for predicting lncRNA-disease associations (2021)
  130. LSTM-based zero-velocity detection for robust inertial navigation (2018)
  131. Predicting aircraft trajectories: a deep generative convolutional recurrent neural networks approach (2018) — link
  132. Deep learning algorithm for data-driven simulation of noisy dynamical system (2019)
  133. Machine learning, a probabilistic perspective (2012)
  134. Multi-object tracking using online metric learning with long short-term memory (2018)
  135. Pattern recognition and machine learning (information science and statistics) (2006)
  136. EM-like learning chaotic dynamics from noisy and partial observations (2019) — link
  137. Attention is all you need (2017)
  138. An EM algorithm for nonlinear state estimation with model uncertainties (2018)
  139. Phase noise estimation for uncoded/coded SISO and MIMO systems (2012) — link
  140. Expectation maximization based parameter estimation by sigma-point and particle smoothing (2014)
  141. Combining recurrent neural networks and adversarial training for human motion synthesis and control (2019)
  142. Looking at the driver/rider in autonomous vehicles to predict take-over readiness (2018) — link
  143. An efficient particle-based online EM algorithm for general state-space models (2015)
  144. BERT: Pre-training of deep bidirectional transformers for language understanding (2018) — link
  145. Reducing BERT pre-training time from 3 days to 76 minutes (2019) — link
  146. Subspace identification for linear systems (1996)
  147. Differential equation and probability inspired graph neural networks for latent variable learning (2022) — link
  148. Parameter estimation for linear dynamical systems (1996) — link
  149. An approach to time series smoothing and forecasting using the EM algorithm (1982)
  150. Bidirectional LSTM-CRF models for sequence tagging (2015) — link
  151. Attention-based CNN-LSTM and XGBoost hybrid model for stock prediction (2022) — link
  152. Mamba: Linear-Time Sequence Modeling with Selective State Spaces (2023) — link
  153. A new approach to linear filtering and prediction problems (1960)
  154. Maximum likelihood estimates of linear dynamic systems (1965)
  155. Maximum likelihood from incomplete data via the EM algorithm (1977)
  156. Unified inference for variational bayesian linear gaussian state-space models (2007)
  157. Finite mixture and markov switching models (2006)
  158. Adam: a method for stochastic optimization (2015)
  159. Dropout: a simple way to prevent neural networks from overfitting (2014)
  160. Graph neural networks and attention-based CNN-LSTM for protein classification (2022) — link
  161. Actions generation from captions (2019) — link
  162. Kalman filter demystified: from intuition to probabilistic graphical model to real case in financial markets (2018) — link
  163. Deep multimodal learning: an effective method for video classification (2018) — link
  164. Predicting blood pressure response to fluid bolus therapy using attention-based neural networks (2018)
  165. Financial series prediction using attention LSTM (2019) — link
  166. A comparison of LSTMs and attention mechanisms for forecasting financial time series (2018) — link
  167. Sequence-to-sequence learning as beam-search optimization (2016)
  168. XGBoost: A Scalable Tree Boosting System (2016)
  169. Sequence to sequence learning with neural networks (2014)
  170. Deep learning (2016)
  171. Long short-term memory (1997)
  172. Epnp: an accurate o ( n) solution to the pnp problem (2009)
  173. Backpropagation through time: what it does and how to do it (1990)
  174. Finding structure in time (1990)
  175. Protein function prediction via graph kernels (2005)

0 of 183 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.