keyword

deep state space models

Deep state space models are hybrid statistical and machine learning architectures that integrate classical state space models with deep neural networks for sequential data analysis and time-series forecasting. In these frameworks, deep neural networks parameterize the latent transition and emission dynamics of the system, replacing or augmenting rigid linear equations with flexible nonlinear mappings learned directly from data. This design enables the model to capture complex temporal dependencies across large and high-dimensional datasets while retaining the core benefits of classical probabilistic state space formulations, including principled uncertainty quantification, data efficiency, and interpretable structural decompositions such as trend and seasonality.

1 item

Time-series forecasting with deep learning: a survey

Bryan Lim, Stefan Zohren

OrganizationsUniversity of Oxford

Why you should read this

Categorizes modern deep learning and hybrid statistical architectures for multi-horizon time series forecasting, providing a clear guide on how different models encode temporal dynamics for operational decision support.

Numerous deep learning architectures have been developed to accommodate the diversity of time series datasets across different domains. In this article, we survey common encoder and decoder designs used in both one-step-ahead and multi-horizon time series forecasting -- describing how temporal information is incorporated into predictions by each model. Next, we highlight recent developments in hybrid deep learning models, which combine well-studied statistical models with neural network components to improve pure methods in either category. Lastly, we outline some ways in which deep learning can also facilitate decision support with time series data.

Source

https://arxiv.org/pdf/2004.13408

Added

2026-09-18