Rough-and-ready assessment of the degree and importance of smoothing in functional estimation

Jones, M. C. (2000). Rough-and-ready assessment of the degree and importance of smoothing in functional estimation. Statistica Neerlandica, 54(1) pp. 37–46.

DOI: https://doi.org/10.1111/1467-9574.00124

Abstract

In nonparametric estimation of functionals of a distribution, it may or may not be desirable, or indeed necessary, to introduce a degree of smoothing into this estimation. In this article, I describe a method for assessing, with just a little thought about the functional of interest, (i) whether smoothing is likely to prove worthwhile, and (ii) if so, roughly how much smoothing is appropriate (in order-of-magnitude terms). This rule-of-thumb is not guaranteed to be accurate nor does it give a complete answer to the smoothing problem. However, I have found it very useful over a number of years; many examples of its use, and limitations, are given.

Viewing alternatives

Metrics

Item Actions

Export

You can export this page using these formats Digital Object Identifier - DOI

About

• Item ORO ID
• 23863
• Item Type
• Journal Item
• ISSN
• 1467-9574
• Keywords
• bandwidth;density derivatives;empirical distribution function;kernel density estimation;kernel functional estimation;mean squared error;smoothed bootstrapping
• Academic Unit or School
• Faculty of Science, Technology, Engineering and Mathematics (STEM) > Mathematics and Statistics
  Faculty of Science, Technology, Engineering and Mathematics (STEM)
• Copyright Holders
• © 2000 VVS
• Depositing User
• Sarah Frain