Ionela-Roxana GLĂVAN, Andreea MIRICĂ, Marian NECULA, Octavian CEBAN, Roxana-Violeta PARTAS-CIOLAN and Liliana CATRINA
Bucharest University of Economic Studies, Romania
The study focuses on building a predictive model that can accurately predict the gender status of victims, identifying the most relevant predictors for the human trafficking cases. The study of this phenomenon is important to evaluate better prediction insights that can effectively be used by the policy makers in the fight against it.
The dataset included in our analysis relies on data from the National Agency against Trafficking in Persons in Romania (ANTIP). The methodology in the present study uses CART popular machine learning technique for classification. Obtained results reveal that the constructed model classifies the victim as female or male with an accuracy of 82,5%.