Core Matrix Regression and Prediction with Regularization

Research output: Chapter in Book/Report/Conference proceedingConference contribution


Many finance time-series analyses often track a matrix of variables at each time and study their co-evolution over a long time. The matrix time series is overly sparse, involves complex interactions among latent matrix factors, and demands advanced models to extract dynamic temporal patterns from these interactions. This paper proposes a Core Matrix Regression with Regularization algorithm (CMRR) to capture spatiotemporal relations in sparse matrix-variate time series. The model decomposes each matrix into three factor matrices of row entities, column entities, and interactions between row entities and column entities, respectively. Subsequently, it applies recurrent neural networks on interaction matrices to extract temporal patterns. Given the sparse matrix, we design an element-wise orthogonal matrix factorization that leverages the Stochastic Gradient Descent (SGD) in a deep learning platform to overcome the challenge of the sparsity and large volume of complex data. The experiment confirms that combining orthogonal matrix factorization with recurrent neural networks is highly effective and outperforms existing graph neural networks and tensor-based time series prediction methods. We apply CMRR in three real-world financial applications: firm earning forecast, predicting firm fundamentals, and firm characteristics, and demonstrate its consistent performance superiority: reducing error by 23%-53% over other state-of-the-art high-dimensional time series prediction algorithms.

Original languageEnglish (US)
Title of host publicationProceedings of the 3rd ACM International Conference on AI in Finance, ICAIF 2022
PublisherAssociation for Computing Machinery, Inc
Number of pages9
ISBN (Electronic)9781450393768
StatePublished - Nov 2 2022
Event3rd ACM International Conference on AI in Finance, ICAIF 2022 - New York, United States
Duration: Nov 2 2022Nov 4 2022

Publication series

NameProceedings of the 3rd ACM International Conference on AI in Finance, ICAIF 2022


Conference3rd ACM International Conference on AI in Finance, ICAIF 2022
Country/TerritoryUnited States
CityNew York

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Finance


  • Tensor algorithm
  • matrix factorization
  • matrix-variate time series prediction
  • recurrent neural networks


Dive into the research topics of 'Core Matrix Regression and Prediction with Regularization'. Together they form a unique fingerprint.

Cite this