Endüstri Mühendisliği Bölümü Koleksiyonu
Permanent URI for this collectionhttps://hdl.handle.net/20.500.11779/1942
Browse
Search Results
Article Citation - WoS: 12Citation - Scopus: 19Minimizing the Misinformation Spread in Social Networks(Taylor and Francis, 2019-11-21) Güney, Evren; Kuban, İ. Kuban Altınel; Tanınmış, Kübra; Aras, Necati; Altinel, I. KubanThe Influence Maximization Problem has been widely studied in recent years, due to rich application areas including marketing. It involves finding k nodes to trigger a spread such that the expected number of influenced nodes is maximized. The problem we address in this study is an extension of the reverse influence maximization problem, i.e., misinformation minimization problem where two players make decisions sequentially in the form of a Stackelberg game. The first player aims to minimize the spread of misinformation whereas the second player aims its maximization. Two algorithms, one greedy heuristic and one matheuristic, are proposed for the first player’s problem. In both of them, the second player’s problem is approximated by Sample Average Approximation, a well-known method for solving two-stage stochastic programming problems, that is augmented with a state-of-the-art algorithm developed for the influence maximization problem.Article Citation - WoS: 6Citation - Scopus: 6A Strong Integer Programming Formulation for Hybrid Flowshop Scheduling(Taylor & Francis, 2019-09-09) Ağralı, Semra; Ünal, A. Tamer; Taşkın, Z. CanerWe consider a hybrid flowshop scheduling problem that includes parallel unrelated discrete machines or batch processing machines in different stages of a production system. The problem is motivated by a bottleneck process within the production system of a transformer producer located in the Netherlands. We develop an integer programming model that minimises the total tardiness of jobs over a finite planning horizon. Our model is applicable to a wide range of production systems organised as hybrid flowshops. We strengthen our integer program by exploiting the special properties of some constraints in our formulation. We develop a decision support system (DSS) based on our proposed optimisation model. We compare the results of our initial optimisation model with an improved formulation as well as with a heuristic that was in use at the company before the implementation of our DSS. Our results show that the improved optimisation model significantly outperforms the heuristic and the initial optimisation model in terms of both the solution time and the strength of its linear programming relaxation.Article Citation - WoS: 2Citation - Scopus: 5Increasing Procurement Efficiency Through Optimal E-Commerce Enablement Scheduling(Emerald Group Publishing Ltd., 2019-06-03) Özlük, Özgür; Cholette, Susan; Clark, Andrew GPurpose: This study aims to show how cost savings can be achieved through optimizing the scheduling of e-commerce enablements. The University of California is one of the largest, most prestigious public education and research systems in the world, yet diminished state support is driving the search for system-wide cost savings. Design/methodology/approach: This study documents the preparation for and rollout of an e-procurement system across a subset of campuses. A math programing tool was developed for prioritizing the gradual rollout to generate the greatest expected savings subject to resource constraints. Findings: The authors conclude by summarizing the results of the rollout, discussing lessons learned and their benefit to decision-makers at other public institutions. Originality/value: The pilot program comprising three campuses has been predicted to yield $1.2m in savings over a one-year period; additional sensitivity analysis with respect to savings, project timelines and other rollout decisions illustrate the robustness of these findings.Article Citation - WoS: 65Citation - Scopus: 72Carbon Price Forecasting Models Based on Big Data Analytics(Taylor and Francis Ltd., 2019-02-17) Çanakoğlu, Ethem; Ağralı, Semra; Yahşi, MustafaAfter the establishment of the European Union's Emissions Trading System (EU-ETS) carbon pricing attracted many researchers. This paper aims to develop a prediction model that anticipates future carbon prices given a real-world data set. We treat the carbon pricing issue as part of big data analytics to achieve this goal. We apply three fundamental methodologies to characterize the carbon price. First method is the artificial neural network, which mimics the principle of human brain to process relevant data. As a second approach, we apply the decision tree algorithm. This algorithm is structured through making multiple binary decisions, and it is mostly used for classification. We employ two different decision tree algorithms, namely traditional and conditional, to determine the type of decision tree that gives better results in terms of prediction. Finally, we exploit the random forest, which is a more complex algorithm compared to the decision tree. Similar to the decision tree, we test both traditional and conditional random forest algorithms to analyze their performances. We use Brent crude futures, coal, electricity and natural gas prices, and DAX and S&P Clean Energy Index as explanatory variables. We analyze the variables' effects on carbon price forecasting. According to our results, S&P Clean Energy Index is the most influential variable in explaining the changes in carbon price, followed by DAX Index and coal price. Moreover, we conclude that the traditional random forest is the best algorithm based on all indicators. We provide the details of these methods and their comparisons.Article Citation - WoS: 6Citation - Scopus: 6Optimal Keyword Bidding in Search-Based Advertising With Budget Constraint and Stochastic Ad Position(Taylor & Francis, 2019-04-20) Özlük, Özgür; Selçuk, Barış; Küçükaydın, HandeThis paper analyses the search-based advertising problem from an advertiser’s view point, and proposes optimal bid prices for a set of keywords targeted for the advertising campaign. The advertiser aims to maximise its expected potential revenue given a total budget constraint from a search-based advertising campaign. Optimal bid prices are formulated by considering various characteristics of the keywords such that the expected revenue from a keyword is a function of the ad’s position on the search page, and the ad position is a stochastic function of both the bid price and the competitive landscape for that keyword. We explore this problem analytically and numerically in an effort to generate important managerial insights for campaign setters.Article Citation - WoS: 8Citation - Scopus: 8Risk Averse Investment Strategies for a Private Electricity Generating Company in a Carbon Constrained Environment(Taylor & Francis, 2019-04-12) Çanakoğlu, Ethem; Ağralı, Semra; Adıyeke, EsraWe study a private electricity generating company that plans to enter a partially regulated market that operates under an active cap and trade system. There are different types of thermal and renewable power plants that the company considers to invest in over a predetermined planning horizon. Thermal power plants may include a carbon capture and storage technology in order to comply with the carbon limitations. We develop a time-consistent multi-stage stochastic optimization model for this investment problem, where the objective is to minimize the conditional value at risk (CV@R) of the net present value of the profit obtained through the planning horizon. We implement the model for a hypothetical generating company located in Turkey. The results show that the developed model is appropriate for determining risk averse investment strategies for a company that operates under carbon restricted market conditions.