Akademik Veri Yönetim Sistemi
Tezin Türü: Doktora
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Endüstri Mühendisliği Bölümü, Türkiye
Tezin Onay Tarihi: 2016
Öğrenci: AHMET FATİH ORTAKAYA
Danışman: CEM İYİGÜN
Özet:Labor market vision has changed from a stock perspective to a flow perspective in the recent years. Majority of labor markets in many high income countries are specified by these gross flows. Due to these large flows, a major issue arises in matching workers and jobs in labour market which results in coexistence of high number of unfilled vacancies and unemployed people. Different approaches are applicable in the literature to match the right candidate with the right job post. Yet, as far as we know a sophisticated statistical analysis or a procedure for employing a Job Matching Scheme (JMS) for Turkish Employment Agency (TEA) does not exist. The main aim of this thesis study is to develop a statistical approach for designing a JMS by proposing a new Classification Algorithm for Categorical Data with Incremental Feature Selection (CACDIFES) and a Matching Algorithm which consists of a combination of scoring and sorting algorithms by using Independently Weighted Overlap Metric (IWOM). Apart from the studies in the literature, this thesis proposes a new Incremental Feature Selection (IFS) algorithm, an Independently Weighted Value Difference Metric (IWVDM) and a modified version of Overlap Metric (OM) which can be applied to any type of categorical data sets. Algorithms proposed in this thesis are applied to TEA data set and data sets obtained from UCI Machine Learning Repository. Experimental results reveal that our proposed metric is superior to previously introduced ones, and our JMS is able to match all vacant jobs with suitable job seekers.