Predicting the Effectiveness of Wildlife Trade Policies Using Machine Learning Techniques

Abstract

Deepening globalization has made the illegal wildlife trade a growing problem, and this paper uses modern information technologies, such as big data analysis and machine learning, for monitoring and evaluation, which are essential for understanding trade dynamics, predicting trends, and assigning responsibilities. This study uses Klein's comprehensive national power equation to establish a scoring system to select the countries or organizations with the most rights, resources, and interest in the management of illegal wildlife trade as the subject of behavioral implementation. Data-driven methods such as nonlinear regression, ARIMA time series forecasting, and the Random Forest algorithm were then used to demonstrate the relevance of the policies and actions of the study subjects to illegal wildlife trade management. The United States, with the highest score 93.0, 1.8 points higher than the second placed IUCN , was identified as the actor with existing policies and actions that are highly relevant to illegal wildlife trade management.