Corpus ID: 88524447

Bootstrapping Generalization Error Bounds for Time Series

  title={Bootstrapping Generalization Error Bounds for Time Series},
  author={Robert Lunde and C. Shalizi},
  journal={arXiv: Statistics Theory},
  • Robert Lunde, C. Shalizi
  • Published 2017
  • Mathematics
  • arXiv: Statistics Theory
  • We consider the problem of finding confidence intervals for the risk of forecasting the future of a stationary, ergodic stochastic process, using a model estimated from the past of the process. We show that a bootstrap procedure provides valid confidence intervals for the risk, when the data source is sufficiently mixing, and the loss function and the estimator are suitably smooth. Autoregressive (AR(d)) models estimated by least squares obey the necessary regularity conditions, even when mis… CONTINUE READING

    Figures from this paper.

    Synthetic learner: model-free inference on treatments over time
    • 4
    • Highly Influenced
    • PDF


    Publications referenced by this paper.
    The Jackknife and the Bootstrap for General Stationary Observations
    • 1,942
    Nonparametric Time Series Prediction Through Adaptive Model Selection
    • R. Meir
    • Computer Science, Mathematics
    • 2004
    • 101
    • PDF
    Automatic Block-Length Selection for the Dependent Bootstrap
    • 591
    • Highly Influential
    • PDF
    Kernel regression smoothing of time series
    • 181
    Introduction to Empirical Processes and Semiparametric Inference
    • 749
    • Highly Influential
    • PDF
    Nonparametric Risk Bounds for Time-Series Forecasting
    • 17
    • PDF
    Central limit theorems for empirical andU-processes of stationary mixing sequences
    • 140
    The Estimation of the Order of an ARMA Process
    • 472
    Estimation and Asymptotic Inference in the AR-ARCH Model
    • 18
    • PDF