Efficient Estimation of Monotone Boundaries

Let g : [0,1] → [0,1] be a monotone non-decreasing Function, such that 0 ≤ g. ≤g(x) ≤ 1 for some constant g., and let. G be the closure of the set{(x,y)∈ [0,1] x [0,1]: 0 ≤ y ≤ g(x)}. We oonsider the problem of estimating the net G from o sample of i.i.d observations uniformly distributed in G. Tim estimation error is measured in the Hausdorff metric. We propose the estimator which is asymptotically efficient in minimax sense.

• metadata
• relations

Document type	Document de travail (Working Paper)
Publication date	1992
Language	Anglais
Number of pages	19 pages
Collection	STAT Discussion Papers - 9209
Affiliations	Academy of Sciences of Russia - Institute of System analysis UCL - EUEN/STAT - Institut de statistique Academy of Sciences of Russia - Institute for Problems of Information Transmission
Keywords	Monotone boundary ; Free disposal hull ; Efficiency ; Minimax risk ; Estimation of support of a density
Links	http://hdl.handle.net/2078.1/120828[Handle]

See also
	[boreal:47953]	Efficient Estimation of Monotone Boundaries