The European Union Asylum Agency (EUAA), the European Commission (Joint Research Centers) and the University of Catania have introduced a new methodology for predicting asylum applications submitted in the EU.
This methodology – based on machine learning and big data, was published in Nature Scientific Reports and aims to increase the preparedness of EU member states for sudden increases in asylum applications as an attempt to process them quickly, fairly and in line with EU law, TheSchengen.com reports.
DynENet’s new machine learning system can anticipate asylum applications submitted in the European Union up to four weeks in advance by integrating traditional immigration and asylum administrative data. It is also destined to do a better job of predicting asylum than the current ARIMA system.
>> The EU Commission publishes the findings of the new compact on migration and introduces new legal frameworks
The EU has been supporting member states since 2011, initially through the European Asylum Support Office (EASO) and since January 2022 through the new EU Asylum Agency (EUAA).
Equipped with an enhanced mandate to produce operational support for Member States under migration stress, this data provided by DynENet will also help increase their internal preparedness and inform authorities as needed.
>> The European Union signs a contract for a large-scale biometric database to strengthen border security
The forecasting strategy follows these steps:
Prepare a fluctuating net elastic model Perform the model estimation on a dynamic data window Select the best dynamic net model (in terms of mean squared error) Predict the covariates for future periods by exploiting lagging variables Also Apply the estimating model to predict the outcome factor.
“Migration is a complex system, drivers are complex, measurement involves uncertainty, and most migration theories are either insufficiently specific or hardly actionable. As a result, forecasting approaches generally focus on specific migration flows, and results are often inconsistent and difficult to generalize”, as It stated in the report, also noting that it focuses on the number of asylum applications submitted in the country by citizens at the international level.
There is a great need for such a tool, especially amid the migration flows observed in the Federation. According to Frontex, the European border agency, the number of illegal immigrants in the region increased by 85 percent (200,000) in 2021, exceeding pre-pandemic levels when strict entry rules did not apply. More specifically, the number of illegal border crossings increased by 36 percent compared to 2019 and an increase of 57 percent over the previous year.