Projects

Causality and machine learning in econometric models

Mgr. Lukáš Lafférs, PhD.

Members of project
Mgr. Miroslav Štefánik, PhD.

• Basic informations
• Characteristics

• Duration of the project: 01/2023 - 12/2025
  ID number: VEGA 1/0398/23

   

• A growing body of current research deals with addressing causal questions using algorithms of machine learning. This project aims to study causal inference questions within econometric models while leveraging the prediction performance of machine learning algorithms. The project will also make the methods more accessible to practitioners.

  While the team will remain open to examine various problems on the topic, there are some more concrete ideas that could be explored. More concretely, the project will attempt to:

  –study causal structure of econometric models in an algorithmic way,

  –explore diagnostic tools that could inform about how appropriate the model assumptions are and how sensitive the results are to violation of identifying assumptions

  –apply the existing methods to make use of high-dimensional data to study employment prospects of registered Slovak job-seekers and the effects of various programmes on both employment and wages.