Date of Award

5-2017

Document Type

Thesis

Degree Name

Master of Science (MS)

College/School

College of Science and Mathematics

Department/Program

Mathematical Sciences

Thesis Sponsor/Dissertation Chair/Project Chair

Andrada E. Ivanescu

Committee Member

Haiyan Su

Committee Member

Andrew McDougall

Subject(s)

Outliers (Statistics) , Multivariate analysis

Abstract

Several methods for performing the identification of outliers are described when dealing with functional data. The methods studied include prediction intervals for detection of dynamic functional outliers as well as related methods from the functional data literature. A comparison of methods is performed using metrics for dynamic outlier identification. Simulations and applications to environmental studies illustrate the applicability of the methods. Results obtained from simulation and application to real dataset suggest that Dynamic Function-on-Function Regression is a preferable method for detecting dynamic outliers. This method can detect outliers at a very high identification rate. Identification rate of dynamic outliers increases when a large number of curves and large size of outliers is observed.

File Format

PDF

Included in

Mathematics Commons

Share

COinS