Thesis Defense - Merve Özer (MSIE)

Merve Özer - M.Sc. Industrial Engineering

Prof. Dr. Ekrem Duman – Advisor

Date: 14.06.2021

Time: 11.00

Location: This meeting will be held ONLINE. Please send an e-mail to gizem.bakir@ozyegin.edu.tr in order to participate in this defense.

METAHEURISTIC APPROACHES FOR THE GENERALIZED ASSIGNMENT PROBLEM OF AN ONLINE EDUCATION WEBSITE

Thesis Committee:

Prof. Dr. Ekrem Duman, Özyeğin University

Assoc. Prof. Dr. Burcu Balçık, Özyeğin University

Assoc. Prof. Dr. Ali Fuat Alkaya, Marmara University

Abstract:

BinYaprak is a TurkishWIN (Turkish Women’s International Network) initiative that offers role model stories for inspiration, educational content, and real-life stories as well as a networking platform through online and offline events. BinYaprak online education website aims to create ways for content and network aggregation between experts and learners for the ease of knowledge discovery. Experts are the people who join the platform for free in order to share their knowledge and experience with interested learners. Learners join the platform for free to learn new skills, access new networks, and find out about new jobs and opportunities. Thus, this study aims to provide an assignment of a learner to an expert which satisfies both sides as much as possible. An expert can be assigned to more than one learner ensuring each learner is assigned at most one expert subject to each expert’s capacity, thus the problem becomes a generalized assignment problem (GAP) which is known to be NP-Hard. In this study, we present an implementation of a new nature-inspired metaheuristic algorithm called Migrating Birds Optimization (MBO) and a hybrid Simulated Annealing (SA) and Tabu Search (TS) method in order to solve the GAP of BinYaprak online education website. In our computational study, we tested both MBO and hybrid SA/TS on small and large-sized instances of GAP of BinYaprak website. Also, we used Gurobi Python API as a baseline reference against which we can compare our heuristic methods’ performances. Our numerical analyzes show that the hybrid SA/TS outperforms the MBO with regards to solution quality and computational effort, hence hybrid SA/TS can be used in practice to achieve high quality solutions.

Bio: 

Merve Özer received her bachelor’s degree in Management from Istanbul Şehir University and graduated as the highest-ranking student. She has experience studying and working at different institutions. She attended the Erasmus+ Exchange program and spent a semester in the UK to take management classes at Keele University, and then she participated in the Erasmus+ Internship program and worked as an academic coordinator at EF International Language School in Brighton / UK. After graduation, she joined the graduate program at Özyeğin University as a MSc student of Industrial Engineering under the supervision of Professor Ekrem Duman in 2020. Her research interests are operations management, supply chain management, metaheuristic approaches, and optimization.