Seminar of Magdalena Kapelko

16 July, 2019
12:30 pma1:30 pm

Speaker: Magdalena Kapelko (Wrocław University of Economics)

Title: “Modelling Environmental Inefficiency under a Quota System”

Date: Tuesday, July 16, 12:30 a.m.

Localication: CIO Seminar Room (Torretamarit Building)

Abstract. This paper introduces the methodology necessary to evaluate inefficiency of regulated decision making units that operate under quotas through Data Envelopment Analysis (DEA), accounting for both quotas’ restrictions and negative environmental externalities of production. Three technical inefficiency measures are proposed: inefficiency in the production of marketed output, environmental inefficiency, and inefficiency with quotas. It is then shown how to aggregate these measures in order to obtain indicators of overall performance. The new approach is illustrated through a numerical example that uses real data available for the European Union dairy sector. The results show that considerable differences in inefficiencies could be found when quotas restrictions are accounted for in the model than in the model without quota imposition, indicating that not accounting explicitly for quotas when measuring performance in regulated sectors may lead to a not accurate estimation of firms’ technical inefficiency.

Seminar of Justo Puerto

3 July, 2019
11:00 ama12:00 pm

Speaker: Justo Puerto (University of Seville)

Title: “Revisiting some models of data analysis in the hand of location theory”

Date: Wednesday, July 3, 11:00 am.

Localication: Classrooms 0.1 and 0.2 of the CIO (Torretamarit Building)

Abstract. This talk presents connections between localization theory and some basic aspects of data analysis. For this, the problem of locating dimensional structures that replace point servers will be addressed. Among the large number of problems that have been studied, we will focus on locating hyperplanes in finite dimension spaces due to their implications in the adjustment of data sets and in the supervised classification. We will approach two hyperplane localization models that give rise to two models of data analysis: the problem of grouping and adjusting, and the problem of supervised learning on datasets with more than two classes. We will present the models and their properties and analyze the computational results on known databases.