ABSTRACT
This chapter discusses the role of interpretability and explainability of machine learning models in explaining socio-economic phenomena. The ever-growing amount of data presents increasing challenges for the effective use of artificial intelligence (AI) algorithms, and many decisions in institutions and enterprises are currently based, at least partly, on these methods. However, the increasing complexity of these models limits their interpretability, which is crucial considering legal regulations and the social consequences of decisions made using these models. Thus, it is essential to comprehend how AI algorithms function, especially for complex non-linear models. This chapter highlights this problem and focuses on the interpretability and explainability of such models as logistic regression, a decision tree, and the extreme gradient boosting algorithm. We examine the importance of gender and health in older adults’ situation in the labor market with regard to the 2030 Agenda for Sustainable Development, focusing on the following Sustainable Development Goals (SDG): Gender (SDG 5), Health (SDG 3), Education (SDG 4), and Decent Work (SDG 8).
