Tezin Türü: Yüksek Lisans
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Uygulamalı Matematik Enstitüsü, Aktüerya Bilimleri Anabilim Dalı, Türkiye
Tezin Onay Tarihi: 2019
Tezin Dili: İngilizce
Öğrenci: SELMA GÜTMEN
Asıl Danışman (Eş Danışmanlı Tezler İçin): Sevtap Ayşe Kestel
Eş Danışman: Gerhard-Wilhelm Weber
Özet:Laborers’ skills are critical for advancement in the labor market in the economy and, eventually, in the areas of health, personal and social security, fulfillment, life quality and expectation. In this respect, it is essential to monitor needed knowledge and available core skills in the market, as well as to make this knowledge accessible to decision makers from economy, business and educational sectors. In Wielkopolskie Voivodeship (Greater Poland region, Poland), a Professional System has been implemented for many years as a vast and diverse dataset, to inspect requests for professional skills by employer, and to accelerate the flow of information between educational system and different areas of the labor market. The first step of the thesis was to preprocess this big dataset to make it suitable for our studies. The main aim of this thesis study is to mathematically model students’ possible contributions (“promise”) to jobs in terms of professionals skills (Z), with particular interest in the dependence of these contributions on common skills (O), general skills (W), any other economic or social circumstances, as well as time variables. Hence, in our model, the response variable is (Z) and all other aspects are implemented as input variables. Each row vector of our dataset is a pair (i,j) where student i is a student and a j is a job offer i applies to. Our aim is to figure out the relationship between the response variable and input variables by applying Linear Regression (LR), Multivariate Adaptive Regression Splines (MARS) and Artificial Neutral Networks (ANNs), as an AI-kind of methodology. We compare the results of MARS model and ANNs model, by the help of statistical performance criteria and statistical graphs, herewith demonstrating the high competitiveness of our MARS approach. Through this analysis we also comment on how to determine the fulfillment between the demands of employers and decision markers of government and university leadership, especially, in the field of education.