Research on Illegal Wildlife Trade Reduction Strategies Based on Principal Component Analysis and Multiple Linear Regression Model
Download as PDF
DOI: 10.25236/icetmr.2024.006
Author(s)
Muxi Xu, HaoYang, Junhao Jin, Siyu Zhang, Jiaxin Yan
Corresponding Author
Muxi Xu
Abstract
In this paper, we propose a five-year data-driven based project that aims to make a useful contribution to reducing the scale of illegal wildlife trade. First, in order to identify the target customers as well as establish the evaluation index system of the project, we successively searched the database with the keywords of illegal wildlife trade and the United States, and identified 15 secondary indicators by combining with the large language model. After using a data crawler to obtain the relevant indicator data, the data were preprocessed. Then the indicators that passed the test were subjected to dimensionality reduction using principal component analysis, and the indicators that did not pass the test were subjected to dimensionality reduction using correlation analysis and t-SNE method. Correlation analysis was conducted between the dimensionality reduced data and the independent variable data to establish the multiple linear regression model. Finally, the linear regression model was used to predict the effectiveness of the program. Comparison of the results proved that the program was effective.
Keywords
illegal wildlife trade; data-driven; predictive modelling; principal component analysis