We gratefully acknowledge support from
the Simons Foundation and member institutions.

Computation

New submissions

[ total of 4 entries: 1-4 ]
[ showing up to 2000 entries per page: fewer | more ]

New submissions for Thu, 29 Jul 21

[1]  arXiv:2107.13430 [pdf, ps, other]
Title: Kernel Density Estimation by Stagewise Algorithm with a Simple Dictionary
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Computation (stat.CO); Methodology (stat.ME)

This paper studies kernel density estimation by stagewise minimization algorithm with a simple dictionary on U-divergence. We randomly split an i.i.d. sample into the two disjoint sets, one to be used for constructing the kernels in the dictionary and the other for evaluating the estimator, and implement the algorithm. The resulting estimator brings us data-adaptive weighting parameters and bandwidth matrices, and realizes a sparse representation of kernel density estimation. We present the non-asymptotic error bounds of our estimator and confirm its performance by simulations compared with the direct plug-in bandwidth matrices and the reduced set density estimator.

[2]  arXiv:2107.13462 [pdf, other]
Title: MSTL: A Seasonal-Trend Decomposition Algorithm for Time Series with Multiple Seasonal Patterns
Subjects: Applications (stat.AP); Computation (stat.CO)

The decomposition of time series into components is an important task that helps to understand time series and can enable better forecasting. Nowadays, with high sampling rates leading to high-frequency data (such as daily, hourly, or minutely data), many real-world datasets contain time series data that can exhibit multiple seasonal patterns. Although several methods have been proposed to decompose time series better under these circumstances, they are often computationally inefficient or inaccurate. In this study, we propose Multiple Seasonal-Trend decomposition using Loess (MSTL), an extension to the traditional Seasonal-Trend decomposition using Loess (STL) procedure, allowing the decomposition of time series with multiple seasonal patterns. In our evaluation on synthetic and a perturbed real-world time series dataset, compared to other decomposition benchmarks, MSTL demonstrates competitive results with lower computational cost. The implementation of MSTL is available in the R package forecast.

Replacements for Thu, 29 Jul 21

[3]  arXiv:2005.08414 (replaced) [pdf, other]
Title: Unbiased MLMC stochastic gradient-based optimization of Bayesian experimental designs
Comments: major revision, 26 pages, 6 figures
Subjects: Computation (stat.CO); Numerical Analysis (math.NA); Machine Learning (stat.ML)
[4]  arXiv:2101.09747 (replaced) [pdf, ps, other]
Title: Numerical issues in maximum likelihood parameter estimation for Gaussian process interpolation
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Computation (stat.CO)
[ total of 4 entries: 1-4 ]
[ showing up to 2000 entries per page: fewer | more ]

Disable MathJax (What is MathJax?)

Links to: arXiv, form interface, find, stat, recent, 2107, contact, help  (Access key information)