Date of Publication

2022

Document Type

Master's Thesis

Degree Name

Master of Science in Statistics

Subject Categories

Statistics and Probability

College

College of Science

Department/Unit

Mathematics and Statistics Department

Thesis Advisor

Frumencio F. Co

Defense Panel Chair

Rechel G. Arcilla

Defense Panel Member

Jayne Lois San Juan
Shirlee R. Ocampo

Abstract/Summary

K-means clustering algorithm is a commonly-used clustering algorithm with many advantages, such as simple understanding, realizing quickly, and processing large datasets conveniently. Count data often applies to many fields, such as medicine, sociology, and psychology. It is an essential statistical data type. Count data is analyzed using some frequently-used models, such as the Poisson regression and the negative binomial regression models. The negative binomial regression model has the phenomenon of overdispersion, wherein the variance is greater than the mean, that exists in the count data. As a consequence, overdispersion data analysis has become a crucial statistical issue.

This thesis focused on studying the application of K-means clustering and the negative binomial regression model in an overdispersed inbound tourism data of the Philippines from 2009 to 2018. The K-means method was used to cluster 58 countries or regions by purpose of travel in the Philippines. The negative binomial regression model was performed for each cluster to identify the determinants of foreign tourist arrivals in the Philippines.

Results showed that only the pattern of the number of tourist arrivals for holiday purpose had a trend stationarity. The number of tourists for holiday purpose was expected to improve the development of tourism. In addition, influencing factors were found to vary among the different clusters.

Abstract Format

html

Language

English

Format

Electronic

Physical Description

ix, 117 leaves

Keywords

Negative binomial distribution

Upload Full Text

wf_yes

Embargo Period

7-1-2022

Download

Share

COinS