Consistency of the k-th nearest neighbor estimator of the relative risk and its application to injury surveillance.
Slavova-S; Kryscio-R; Bunn-T
Proceedings of the Thirty Seventh Spring Conference of the Union of Bulgarian Mathematicians, Mathematics and Education in Mathematics, April 2-6, 2008, Borovetz, Bolgaria. 2008 Apr; :232-238
The relative risk function is defined as the ratio of two probability density functions, usually cases to controls, at a fixed point. The nonparametric k-th nearest neighbor (kNN) approach is used for the density estimation. The kNN relative risk estimator at a fixed location is shown to be asymptotically consistent. An application of the kNN relative risk estimator to injury surveillance is discussed.
Mathematical-models; Risk-factors; Beta-radiation; Sampling; Diseases; Injuries; Demographic-characteristics; Divers; Truck-drivers; Trucking; Age-groups; Injury-prevention; Education; Motor-vehicles; Humans; Men; Women
Svetla Slavova, Department of Statistics, University of Kentucky, 817 Patterson Office Tower, Lexington, KY 40506
Proceedings of the Thirty Seventh Spring Conference of the Union of Bulgarian Mathematicians, Mathematics and Education in Mathematics, April 2-6, 2008, Borovetz, Bolgaria
University of Kentucky