Doctoral seminar "Analysis of the labor market using machine-learning" [19.12.2022]

We kindly invite you to join the last doctoral seminar in this calendar year. During the meeting Maciej Nasiński, M.A. will present his research.

The meeting will be held on December 19th 2022 at 3 p.m. in hybrid mode: on-site in room A409 at the Faculty of Economic Sciences and via Zoom platform.

Link to the meeting: https://uw-edu-pl.zoom.us/j/91533433049?pwd=OUNzYzdXUjZsYmRGM1pRSWdMbmYvQT09

Meeting ID: 915 3343 3049

Access code: 400347

If you would like to take part in the seminars or present your research, please contact prof. Jan Jakub Michałek or Piotr Mandera.

Abstract:

The Labour Market and connected with its problems are socially essential topics. The main objective is the application of machine-learning tools to improve the estimation quality and expand the research area for the Labour Market. My series of papers is more than implementing existing solutions, as new solutions are proposed. The newly invented cat2cat procedure helps to unify an inconsistently coded categorical variable in a panel dataset. The cat2cat procedure extends the scope of the available statistical analyses in a panel dataset with inconsistently coded categorical variables, which are ordinarily removed or force dataset aggregation. The cat2cat procedure was implemented for the occupational variable (The International Standard Classification of Occupations), which was later used in the regression model for the panel dataset without losing degrees of freedom and assessment quality. Another application of machine-learning tools is estimating the impact of minimum wages. The result of classification models enables the probabilistic representation of being exposed to the minimum wage, which allows the inclusion of the unemployed to the analysis.