Regularized estimation in sparse high-dimensional time series models

Abstract

Many scientific and economic problems involve the analysis of high-dimensional time series datasets. However, theoretical studies in high-dimensional statistics to date rely primarily on the assumption of independent and identically distributed (i.i.d.) samples. In this work, we focus on stable Gaussian processes and investigate the theoretical properties of ${\ell }_1$-regularized estimates in two important statistical problems in the context of high-dimensional time series: (a) stochastic regression with serially correlated errors and (b) transition matrix estimation in vector autoregressive (VAR) models. We derive nonasymptotic upper bounds on the estimation errors of the regularized estimates and establish that consistent estimation under high-dimensional scaling is possible via ${\ell }_1$-regularization for a large class of stable processes under sparsity constraints. A key technical contribution of the work is to introduce a measure of stability for stationary processes using their spectral properties that provides insight into the effect of dependence on the accuracy of the regularized estimates. With this proposed stability measure, we establish some useful deviation bounds for dependent data, which can be used to study several important regularized estimates in a time series setting.

Published at http://dx.doi.org/10.1214/15-AOS1315 in the Annals of Statistics (http://www.imstat.org/aos/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (5)

• Nearly unbiased variable selection under minimax concave penalty
• Covariance regularization by thresholding
• Covariance and precision matrix estimation for high-dimensional time series
• Thresholded Lasso for high dimensional variable selection and statistical estimation
• Asymptotic equivalence of regularization methods in thresholded parameter space

Cited by in corpus (33)

• Neural Granger Causality
• Signal Processing on Graphs: Causal Modeling of Unstructured Data
• Granger Causality in Multi-variate Time Series using a Time Ordered Restricted Vector Autoregressive Model
• Low Rank and Structured Modeling of High-dimensional Vector Autoregressions
• Networks, Dynamic Factors, and the Volatility Analysis of High-Dimensional Financial Series
• Evaluation of Granger causality measures for constructing networks from multivariate time series
• Testing for high-dimensional white noise using maximum cross-correlations
• FNETS: Factor-adjusted network estimation and forecasting for high-dimensional time series
• Regularized estimation of linear functionals of precision matrices for high-dimensional time series
• LAVARNET: Neural Network Modeling of Causal Variable Relationships for Multivariate Time Series Forecasting
• Estimation of Large Covariance and Precision Matrices from Temporally Dependent Observations
• An autocovariance-based learning framework for high-dimensional functional time series
• SILVar: Single Index Latent Variable Models
• Penalized Estimation and Forecasting of Multiple Subject Intensive Longitudinal Data
• Variable Screening for High Dimensional Time Series
• Sampling Requirements for Stable Autoregressive Estimation
• High-Dimensional Multivariate Time Series With Additional Structure
• Non-Asymptotic Guarantees for Reliable Identification of Granger Causality via the LASSO
• Regularization Approach for Network Modeling of German Power Derivative Market
• Estimation of dynamic networks for high-dimensional nonstationary time series
• Predictive Quantile Regression with Mixed Roots and Increasing Dimensions: The ALQR Approach
• Finite Sample Theory for High-Dimensional Functional/Scalar Time Series with Applications
• Vector Autoregressive Models with Spatially Structured Coefficients for Time Series on a Spatial Grid
• Local Projection Inference in High Dimensions
• Correlating Time Series with Interpretable Convolutional Kernels
• A Fast Detection Method of Break Points in Effective Connectivity Networks
• On Conditional Independence Graph Learning From Multi-Attribute Gaussian Dependent Time Series
• Factor-guided estimation of large covariance matrix function with conditional functional sparsity
• Factor-Driven Network Informed Restricted Vector Autoregression
• Penalised inference for autoregressive moving average models with time-dependent predictors
• Causal Discovery in High-Dimensional Point Process Networks with Hidden Nodes
• Two step estimations via the Dantzig selector for models of stochastic processes with high-dimensional parameters
• Sparse estimation of parameter support sets for generalized vector autoregressions by resampling and model aggregation