Research map / Architectures and efficiency / RNNs and normalization
Compact Autoregressive Network
Di Wang, Feiqing Huang, Jingyu Zhao, Guodong Li, Guangjian Tian
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use Tucker-factorized TAR-2 instead of RNN | US macroeconomic forecasting · US macroeconomic dataset, rolling forecast | RMSE: 0.9217 -> 0.8958 | Improved |
| Add a second parallel lane to TAR (ablation) | TAR · US macroeconomic dataset, rolling forecast | RMSE: 0.9104 -> 0.8958 | Improved |
| Add nonlinear activations to LTAR (ablation) | LTAR · NL-DGP, T=500 | RMSE: 1.0126 -> 1.0110 | Improved |
| Add nonlinear activations to LTAR (ablation) | LTAR · L-DGP, T=100 | RMSE: 1.0388 -> 1.0998 | Hurt |
| Use two parallel lanes instead of one (ablation) | TAR · L-DGP, T=100 | RMSE: 1.0998 -> 1.1238 | Hurt |
| Use two parallel lanes instead of one (ablation) | TAR · L-DGP, T=500 | RMSE: 1.0585 -> 1.0493 | Improved |
| Use two parallel lanes instead of one (ablation) | TAR · NL-DGP | RMSE: 1.0712 -> 1.0703; 1.0110 -> 1.0111 | Mixed |
| Use Tucker structure over ordinary least squares | Linear autoregressive network · Linear autoregressive estimation | sample complexity: N^2P/ε^2; (r_1r_2r_3+Nr_1+Nr_2+Pr_3)/ε^2 | Improved |
Closest papers by meaning
- BARNN: A Bayesian Autoregressive and Recurrent Neural Network (ICML 2025 · similarity 0.86)
- DeltaProduct: Improving State-Tracking in Linear RNNs via Householder Products (NeurIPS 2025 · similarity 0.84)
- Autoregressive Learning in Joint KL: Sharp Oracle Bounds and Lower Bounds (NeurIPS 2026 · similarity 0.84)
- TNPAR: Topological Neural Poisson Auto-Regressive Model for Learning Granger Causal Structure from Event Sequences (AAAI 2024 · similarity 0.83)
- TStruct: Learning Shared Temporal Structures for Long-Term Time-Series Forecasting (NeurIPS 2026 · similarity 0.83)
- Structured Linear CDEs: Maximally Expressive and Parallel-in-Time Sequence Models (NeurIPS 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
