We propose interval estimates for the point(s) of maximum of the intensity function of an inhomogeneous Poisson process. Possible applications include the identification of times when customer arrivals or accident occurrences are typically maximal. In particular we present an application where the goal has been to identify time periods of maximal ship arrival intensity at Keelung harbor. Asymptotic results c~ncerning the coverage probability are given under minimal smoothness assumptions and without assuming a unique point of global maximum. The proposed confidence sets may also be used in the context of mode estimation for probability densities. The finite sample performance is investigated in a small simulation study.
2000年國際數學與統計研討會暨第34屆中華民國數學會年會：慶祝淡江大學創校五十週年校慶 (2000 international conference on mathematics and statistics and 34th annual meeting of mathematical society of R.O.C. : commemorating the 50th anniversary of Tamkang University), pp.31