Year: 2023
Type: article
Abstract: Models of human migration predictions are adequately important to Governments and society as comprehensive tools for making predictions on population changes in the near future.The prediction models would accommodate a database of the readiness and effect in labor markets, the risk of spreading infectious diseases, and the capacity in enduring economic fluctuations.However, as migrations are complex and unpredictable in the huge demographic processes, predicting migration can be notorious and have a high error rate.Using a machine learning approach that can intelligently predict the reverse migration based on the tested data set is a significant way to reduce errors.Thus, this paper aims to explore and compare machine learning models namely three gradient-boosted trees, a random forest, and a decision tree in reverse migration forecasting.Besides the performance of the comparisons, this paper presents the specific weight correlation in each of the machine learning models to describe the importance of the reverse migration factors to the model.
Institution: Universiti Teknologi MARA
Language: English
Topic: Migration and Labor Dynamics
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: diamond
Funder: Universiti Teknologi MARA
Award: G854476752
OpenAlex ID: W4322496172
