Date of Award

1-1-2017

Language

English

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

College/School/Department

Department of Epidemiology and Biostatistics

Program

Biostatistics

Content Description

1 online resource (ii, ix, 210 pages)

Dissertation/Thesis Chair

Igor G. Zurbenko

Committee Members

Gregory A. DiRienzo, Robert F. Henry, Daniel M. Fernandez, Yucel M. Recai

Keywords

El Niño-like component, Frequency domain analysis, Global surface humidity, Reconstruction, Spatial wave separation, Spatiotemporal data analysis, Fourier transformations, Atmospheric waves, Atmospheric temperature, Boundary layer (Meteorology), Earth temperature, Climatic changes, Nonparametric statistics

Subject Categories

Biostatistics | Climate

Abstract

Unlike one-dimensional wave reconstruction, reconstruction 2D spatial wave via Fourier Transform doesn’t look like a non-parametric algorithm. In other words, we need the wave frequency and wave direction information to recover the spatial wave via Fourier Transform, especially when the stress of noise is present. The direct consequence is that accurate estimations of wave parameters are need for reconstructing of spatial waves. To this end, we propose to improve the accuracy of motion image scale detection and parameter estimations with optimization based on Kolmogorov-Zurbenko periodogram (KZP) information. Related methods and algorithms are denoted under the name of Kolmogorov-Zurbenko wave separations. The non-parametric approach of reconstruction via Kolmogorov-Zurbenko Filters (KZ Filters) is also included in this range.

Download

Share

COinS