Learning the smoothness of noisy curves with application to online curve estimation
arXiv:2009.03652 · doi:10.1214/22-EJS1997
Abstract
Combining information both within and across trajectories, we propose a simple estimator for the local regularity of the trajectories of a stochastic process. Independent trajectories are measured with errors at randomly sampled time points. Non-asymptotic bounds for the concentration of the estimator are derived. Given the estimate of the local regularity, we build a nearly optimal local polynomial smoother from the curves from a new, possibly very large sample of noisy trajectories. We derive non-asymptotic pointwise risk bounds uniformly over the new set of curves. Our estimates perform well in simulations. Real data sets illustrate the effectiveness of the new approaches.
References in corpus (13)
- Estimators of Fractal Dimension: Assessing the Roughness of Time Series and Spatial Data
- Bandwidth selection in kernel density estimation: Oracle inequalities and adaptive minimax optimality
- Confidence bands in density estimation
- Optimal estimation of the mean function based on discretely sampled functional data: Phase transition
- Review of Functional Data Analysis
- Model-Predictive Policy Learning with Uncertainty Regularization for Driving in Dense Traffic
- Sharp asymptotics of the functional quantization problem for Gaussian processes
- Adaptive Bernstein-von Mises theorems in Gaussian white noise
- Estimation of the Hurst parameter from discrete noisy data
- Estimation of fractal dimension for a class of Non-Gaussian stationary processes and fields
- Monash University, UEA, UCR Time Series Extrinsic Regression Archive
- Testing the regularity of a smooth signal
- A Framework for Probabilistic Generic Traffic Scene Prediction