Prediction of Irregular Bangladesh-EU Migration Trends Using Machine Learning Techniques

Remittances are important economic contributors to a countryâs economy. It has increased national savings, lowered foreign exchange limits, and helped balance payments and development budgets. Migrant workers are the main source of these remittance revenues. Many workers worldwide leave their homes and move to different parts of the world for overseas employment. These workers are authorized and have documentation of their country of origin and final destination. Besides, many workers have migrated illegally globally, particularly to European Union (EU) countries via marine routes. However, illegal immigration leads to a loss of national revenue due to non-payment of taxes and social security contributions and problems with living circumstances and legal issues. Therefore, this paper aims to present a time series forecast of irregular migrants in EU countries from Bangladesh over the next five years. We used five methods to analyze this forecasting: ARIMA (an autoregressive integrated moving average), XGBoost Series, Decision Tree, Feed Forward Neural Network, and CatBoost Regressor. In addition, we used data from the European Border and Coast Guard Agency to perform the time series forecast. A field intervention is carried out to assess the gender, age, education, socioeconomic level, and skill profile of migrant workers. Moreover, a few case studies of migrant workers and a survey were conducted to investigate the underlying causes of irregular migration. Finally, this study explains the underlying cause of irregular migration and its repercussions. It can provide some ideas for decreasing the loss of human life and property by reducing irregular migration to help take essential actions in the home and destination countries.