Date of Publication
2023
Document Type
Bachelor's Thesis
Degree Name
Bachelor of Science in Statistics Major in Actuarial Science
Subject Categories
Statistics and Probability
College
College of Science
Department/Unit
Mathematics and Statistics Department
Thesis Advisor
Regina M. Tresvalles
Defense Panel Chair
Rechel G. Arcilla
Defense Panel Member
Vio Jianu C. Mojica
Abstract/Summary
The inequality in income distribution among Filipino families generates a substantial problem that is ineffectively resolved by administered economic programs in the past. In this research, Recursive-Feature Elimination, Cross-Validated was implemented to obtain the optimal number of features for Random Forest Classification in order to acquire the most relevant features from the 2018 Family Income and Expenditure Survey data that influence family income class. Examining the precision, recall, and f1-score, the results were 25 relevant features from the 111 features, as well as an increase in the classification performance of the family income classes. Moreover, the analysis determined that the demographic age of household head, number of family members, household assets such as house floor area, number of bedrooms, radios, televisions, and refrigerators, as well as expenditures such as the total food and non-food expenditures, are the relevant features that influence family income class.
Abstract Format
html
Language
English
Format
Electronic
Keywords
Income
Recommended Citation
Angue, K. M., Cruz, K. M., & Hilomen, J. C. (2023). Feature selection of the determining factors of family income class using FIES results 2018: A random forest approach with recursive feature elimination, cross-validated. Retrieved from https://animorepository.dlsu.edu.ph/etdb_math/25
Upload Full Text
wf_yes
Embargo Period
4-19-2023
