Author

Broti Garai

Date of Graduation

2016

Document Type

Dissertation

Degree Type

PhD

College

Statler College of Engineering and Mineral Resources

Department

Lane Department of Computer Science and Electrical Engineering

Committee Chair

Robert Mnatsakanov

Committee Co-Chair

Krzysztof Chris Ciesielski

Committee Member

James E Harner

Committee Member

Erin Leatherman

Committee Member

Michael Mays

Abstract

In this research, we propose formulas for recovering the regression function based on the product moments as well as recovering the distributions and derivative function in some indirect models. The upper bounds for the uniform rate of approximations are also derived. For regression functions two cases where the support of underlying functions is bounded and unbounded from above are studied. Based on the proposed approximations, new types of nonparametric estimates of the cumulative distribution, the density functions and the derivative functions in multiplicative-censoring model, as well as the approximations of conditional variance, are constructed. The suggested approach is also applied in the demixing problem and the constrained deconvolution problem, as well as for recovery of a distribution for unknown finite support. The absolute and mean squared errors of corresponding estimates are investigated as well. A simulation study justifies the consistency of the proposed estimates.

Share

COinS