The Development of Some Stochastic Models of Mathematical Sciences
Abstract
The research accomplishments under the present contract fall in three areas of mathematical sciences: nonparametric probability estimation, environmental problems, and reliability modelings. Utilizing the theory of sample characteristics functions and the Fourier representation of the kernel estimator, we have derived rates of convergence for kernel estimators in a large class of Hilbert spaces. Preliminary results have been obtained on two unresolved problems in the area of penalized maximum-likelihood estimators. The first problem we dealt with is developing the consistency conditions of such estimators. The second accomplishment was the development of an objective criterion for choosing a smoothness parameter. (kr)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1980
- Accession Number
- ADA215170
Entities
People
- Chris P. Tsokos
Organizations
- University of South Florida