Predicting unemployment risk: a study for MAITT (machine learning and AI supported services)

As part of MAITT, a 2019–2022 project commissioned by the Estonian Research Council to test machine learning in public services, our team modelled the probability that employed people become unemployed or drop out of work, examined the factors behind that risk, and evaluated the impact of the Estonian Unemployment Insurance Fund’s services for employed people.

Read the final report here. Our analysts Andres Võrk and Taavi Unt were key contributors in section 4.1: predicting unemployment risk.