mlsanityData to 14 Oct 2026

Time-varying Factor Augmented Vector Autoregression with Grouped Sparse Autoencoder

arXiv 2503.04386 ↗First posted 6 Mar 2025stat.ML cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Recent economic events, including the global financial crisis and COVID-19 pandemic, have exposed limitations in linear Factor Augmented Vector Autoregressive (FAVAR) models for forecasting and structural analysis. Nonlinear dimension techniques, particularly autoencoders, have emerged as promising alternatives in a FAVAR framework, but challenges remain in identifiability, interpretability, and integration with traditional nonlinear time series methods. We address these challenges through two contributions. First, we introduce a Grouped Sparse autoencoder that employs the Spike-and-Slab Lasso prior, with parameters under this prior being shared across variables of the same economic category, thereby achieving semi-identifiability and enhancing model interpretability. Second, we incorporate time-varying parameters into the VAR component to better capture evolving economic dynamics. Our empirical application to the US economy demonstrates that the Grouped Sparse autoencoder produces more interpretable factors through its parsimonious structure; and its combination with time-varying parameter VAR shows superior performance in both point and density forecasting. Impulse response analysis reveals that monetary policy shocks during recessions generate more moderate responses with higher uncertainty compared to expansionary periods.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 15 of its 53 references that are in the map. First posted on arXiv 6 Mar 2025.

Builds on

15 of 53 references.

References · 15

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2013Auto-Encoding Variational Bayescited by 4,053
  3. 2014NICE: Non-linear Independent Components Estimationcited by 1,058
  4. 2018Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representationscited by 941
  5. 2019Variational Autoencoders and Nonlinear ICA: A Unifying Frameworkcited by 503
  6. 2021Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICAcited by 160
  7. 2020Bayesian Neural Networks: An Introduction and Surveycited by 131
  8. 2023Additive Decoders for Latent Variables Identification and Cartesian-Product Extrapolationcited by 58
  9. 2021Identifiable Deep Generative Models via Sparse Decodingcited by 54
  10. 2014Bayesian linear regression with sparse priorscited by 40
  11. 2017Adaptive Estimation in Structured Factor Models with Applications to Overlapping Clusteringcited by 19
  12. 2020Spike-and-Slab Meets LASSO: A Review of the Spike-and-Slab LASSOcited by 10
3 more
  1. 2020Deep Dynamic Factor Modelscited by 6
  2. 2018A fast algorithm with minimax optimal guarantees for topic models with an unknown number of topicscited by 5
  3. 2023Non-linear dimension reduction in factor-augmented vector autoregressionscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Dissecting Chronos: Sparse Autoencoders Reveal Causal Feature Hierarchies in Time Series Foundation Models
  2. 2024STORM: A Spatio-Temporal Factor Model Based on Dual Vector Quantized Variational Autoencoders for Financial Trading
  3. 2026TimeSAE: Causal Sparse Decoding for Faithful Explanations of Black-Box Time Series Models
  4. 2026Beyond Coefficients: Forecast-Necessity Testing for Interpretable Causal Discovery in Nonlinear Time-Series Models
  5. 2024Sparse-VQ Transformer: An FFN-Free Framework with Vector Quantization for Enhanced Time Series Forecasting
  6. 2026TSAE: Structured Sparse Autoencoders for Interpreting Time-Series Forecasting Models
  7. 2026SPaRSe-TIME: Saliency-Projected Low-Rank Temporal Modeling for Efficient and Interpretable Time Series Prediction
  8. 2026From Causal Discovery to Dynamic Causal Inference in Neural Time Series