Researchers have developed a new non-asymptotic analysis for estimating vector autoregressive models, particularly for systems with heavy-tailed noise. The study establishes sample complexity bounds, showing that estimation error is influenced by noise dimension and the number of samples. This approach is generalized to various noise distributions, including sub-exponential and sub-Gaussian, and is applied to autoregressive models with exogenous inputs, demonstrating that the dimension factor is independent of model order. AI
IMPACT This research advances theoretical understanding in statistical learning, potentially improving the robustness of models dealing with noisy data.
RANK_REASON Academic paper published on arXiv detailing a new analysis method for machine learning. [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- arXiv
- Autoregressive models with exogenous inputs
- CatalyzeX
- DagsHub
- Gotit.pub
- Hugging Face
- IArxiv
- Least Squares Estimation and Adaptive Prediction in Non-Linear Stochastic Regression Models with Applications to Time Series and Stochastic Systems
- ScienceCast
- Sub-exponential noise
- Sub-Gaussian noise
- Vector Autoregressive Models and Granger Causality in Time Series Analysis in Nursing Research: Dynamic Changes Among Vital Signs Prior to Cardiorespiratory Instability Events as an Example
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
