Date

2015

Document Type

Dissertation

Degree

Doctor of Philosophy

Department

Mathematics

First Adviser

Huang, Wei-Min

Other advisers/committee members

Eisenberg, Bennett; Wu, Ping-Shi; Magent, Michael

Abstract

This dissertation aims at developing an efficient and powerful method for comparing time series in different applications. We introduce a large sample nonparametric test for the hypothesis of equal stochastic processes of two independent time series. The proposed method is based on a modified version of Sample Entropy (SampEnM). Under certain conditions, we establish the validity of the asymptotic distribution of the proposed test statistic and derive its standard error. As an alternative, we also show that it is possible to bootstrap the distribution of the test statistic under the null and obtain bootstrap standard error and critical values. Applications to various ARMA models and real-life data imply that the proposed method is not only simpler to compute but may also outperform existing competitor methods.

Available for download on Monday, January 15, 2018

Included in

Mathematics Commons

Share

COinS