Ankara, Turkey

TOBB University of Economics and Technology is a private non-profit foundation university in Ankara, Turkey. Wikipedia.

Time filter

Source Type

Kuyzu G.,TOBB University of Economics and Technology
Computers and Operations Research | Year: 2017

In collaborative truckload transportation procurement, the collaborating shippers aim to jointly identify and submit tours with little or no asset repositioning to a carrier, as opposed to submitting individual lanes, in return for more favorable rates. In order to maximize savings, the shippers must solve a Lane Covering Problem (LCP), which in more mathematical terms means to cover a subset of arcs in a directed graph by a set of constrained cycles with minimum total cost. Previous works in the literature have proposed effective greedy algorithms or the solution of the NP-Hard LCP variants. This paper incorporates a new constraint into the LCP, motivated by the need to limit the number of partners with whom the collaborative tours must be coordinated. An integer programming model is formulated, and column generation and branch-and-price approaches are developed for the solution of the resulting LCP variant. © 2016 Elsevier Ltd

Agency: European Commission | Branch: FP7 | Program: CP-IP | Phase: AAT.2011.4.4-4. | Award Amount: 37.57M | Year: 2011

The ESPOSA project will develop and integrate novel design and manufacture technologies for a range of small gas turbine engines up to approx. 1000 kW to provide aircraft manufacturers with better choice of modern propulsion units. It will also deal with engine related systems which contribute to the overall propulsion unit efficiency, safety and pilot workload reduction. Through the newly developed design tools and methodologies for the engine/aircraft integration the project will also contribute to the improved readiness for new turbine engines installation into aircraft. New technologies and knowledge gained through the ESPOSA project will provide European general aviation industry with substantially improved ability to develop and use affordable and environmentally acceptable propulsion units and reliable aircraft systems minimizing operating costs, while increasing the level of safety. The new engine systems and engine technologies gained from ESPOSA should deliver 10-14% reduction in direct operating costs (DOC) and reduce significantly the pilot workload. The ESPOSA project is oriented on turbine engine technologies tailored for a small aircraft up to 19 seats (under CS-23/FAR23) operated on the scheduled and non-scheduled flights. The research work comprises performance improvements of key engine components, their improved manufacture in terms of costs and quality. New engine component technologies will be backed by novel modern electronic engine control based on COTS, pioneering the engine health monitoring for small engines and providing new more electric solutions for fuel and propeller control systems. Project activities will include extensive validation on the test rigs. The most appropriate technologies according to value/cost benefit will be selected and integrated into functional complexes and further evaluated on the engine test beds. The functionality of certain project outcomes will also be demonstrated and validated in-flight conditions.

Yalta A.T.,TOBB University of Economics and Technology
Energy Economics | Year: 2011

We employ a maximum entropy bootstrap based framework to analyze the energy consumption and real GDP nexus between 1950 and 2006 in Turkey. Our approach provides more accurate inference in comparison to conventional hypothesis tests based on asymptotic theory. It also avoids preliminary testing and shape-destroying transformations such as differencing and detrending. The bivariate analysis as well as a multivariate framework controlling for exchange rate and oil prices shows no evidence of a causal relation. Our results are robust to both the number of lags and the time period chosen. We also perform a cointegration analysis of the data and point out a common misunderstanding in the literature regarding the concept of causation. © 2010 Elsevier B.V.

Gultekin H.,TOBB University of Economics and Technology
International Journal of Production Economics | Year: 2012

This study considers the throughput optimization in a two-machine flowshop producing identical jobs. Unlike the general trend in the scheduling literature, the machines are assumed to be capable of performing different operations. As a consequence, one of the three operations that a job requires can only be processed by the first and another operation can only be processed by the second machine. These are called fixed operations. The remaining one is called the flexible operation and can be processed by any one of the machines. The machines are assumed to have different technological properties, i.e. non-identical, so that the processing time of the flexible operation has different values on the two machines. We first consider the problem of assigning the flexible operations to the machines for each job in order to maximize the throughput rate. We develop constant time solution algorithms for infinite and zero capacity buffer spaces in between the machines. We then analyze the benefits of flexibility. Managerial insights are provided regarding the changes in the makespan as well as the associated cost with respect to the increase in the level of flexibility. © 2012 Elsevier B.V. All rights reserved.

Bagci G.B.,TOBB University of Economics and Technology
Physica A: Statistical Mechanics and its Applications | Year: 2015

Absrtact Bento et al. (2015) recently proposed to use the third law of thermodynamics as a test for the generalized entropies, since the third law should be respected by any statistical mechanical entropy regardless of the explicit form of the Hamiltonian. We first consider the Rényi entropy which is the sole additive generalized entropy, and show that it violates the third law for q>1. Interestingly, this is exactly the interval where the Rényi entropy is neither concave nor convex. We then consider the homogeneous entropy and show that it also violates the third law for 0

Ubeyli E.D.,TOBB University of Economics and Technology
Expert Systems with Applications | Year: 2010

The aim of the study is classification of the electroencephalogram (EEG) signals by combination of the model-based methods and the least squares support vector machines (LS-SVMs). The LS-SVMs were implemented for classification of two types of EEG signals (set A - EEG signals recorded from healthy volunteers with eyes open and set E - EEG signals recorded from epilepsy patients during epileptic seizures). In order to extract the features representing the EEG signals, the spectral analysis of the EEG signals was performed by using the three model-based methods (Burg autoregressive - AR, moving average - MA, least squares modified Yule-Walker autoregressive moving average - ARMA methods). The present research demonstrated that the Burg AR coefficients are the features which well represent the EEG signals and the LS-SVM trained on these features achieved high classification accuracies. © 2009 Elsevier Ltd. All rights reserved.

Ubeyli E.D.,TOBB University of Economics and Technology
Expert Systems with Applications | Year: 2010

A new approach based on the implementation of probabilistic neural network (PNN) is presented for classification of electroencephalogram (EEG) signals. In practical applications of pattern recognition, there are often diverse features extracted from raw data which needs recognizing. Because of the importance of making the right decision, the present work is carried out for searching better classification procedures for the EEG signals. Decision making was performed in two stages: computation of Lyapunov exponents as feature vectors and classification using the classifiers trained on the extracted features. The aim of the study is classification of the EEG signals by the combination of Lyapunov exponents and the PNN. The purpose is to determine an optimum classification scheme for this problem and also to infer clues about the extracted features. The present research demonstrated that the Lyapunov exponents are the features which well represent the EEG signals and the PNN trained on these features achieved high classification accuracies. © 2009 Elsevier Ltd. All rights reserved.

Derya Ubeyli E.,TOBB University of Economics and Technology
Expert Systems with Applications | Year: 2010

An approach based on the consideration that electrocardiogram (ECG) signals are chaotic signals was presented for automated diagnosis of electrocardiographic changes. This consideration was tested successfully using the nonlinear dynamics tools, like the computation of Lyapunov exponents. Recurrent neural network (RNN) was implemented and used as basis for detection of variabilities of ECG signals. Four types of ECG beats (normal beat, congestive heart failure beat, ventricular tachyarrhythmia beat, atrial fibrillation beat) obtained from the PhysioBank database were classified. Decision making was performed in two stages: computing features which were then input into the RNN and classification using the RNN trained with the Levenberg-Marquardt algorithm. The research demonstrated that the Lyapunov exponents are the features which are well representing the ECG signals and the RNN trained on these features achieved high classification accuracies. © 2009 Elsevier Ltd. All rights reserved.

Agency: European Commission | Branch: FP7 | Program: MC-CIG | Phase: FP7-PEOPLE-2011-CIG | Award Amount: 100.00K | Year: 2011

We propose to develop scalable solvers for integral equation based nonlocal (NL) problems such as peridynamics (PD). Heterogeneity will also be studied due to utmost importance of composite materials to numerous applications in material science and structural mechanics. Robustness of the solvers with respect to heterogeneity and multiscale finite element discretizations are the subsequent directions to pursue. Since the impact of nonlocality on solvers has never been studied before, this research initiative is unique, transformative, and has great potential to create a solver subfield: nonlocal domain decomposition methods (DDM). We propose to study both the algorithmic and theoretical aspects of DDM. The solver research has the potential to reveal multiscale implications associated to NL modeling. We recently proved fundamental conditioning results indicating that the weak formulation of PD can be bounded independently of the mesh size, meaning that one can increase the resolution without increasing the condition number. Scalable and robust solver technologies will create a great impact on simulation capabilities of nonlocal problems at large. In particular, PD will be used for more complex and realistic NL applications because scalable solvers will directly impact the modeling and simulation capability. There is also imminent need for robust preconditioning in the computational material science community as composite materials become industry standard. For instance, Airbus heavily uses light weight composite materials in modern aircrafts.

Agency: European Commission | Branch: FP7 | Program: MC-CIG | Phase: FP7-PEOPLE-2011-CIG | Award Amount: 75.00K | Year: 2011

Numerous industries use revenue management (RM) to forecast demand for products and to determine product prices and availability. Airline and hotel companies, who spearheaded RM developments in the 1990s, have reported impressive annual revenue gains: American Airlines realized $500 million and Marriott realized $100 million. Today, however, the $218 billion airline and hotel sectors struggle to maintain protability in a marketplace dominated by online purchases. Traditional RM systems have struggled to adapt to these new market conditions, leading to calls for fundamentally new choice-based RM systems that use discrete choice models to forecast demand in a way that better reects todays purchasing environment. Choice based revenue management has the potential to revolutionize the way that companies determine their pricing and revenue strategies. This approach incorporates consumer behavior into classical revenue management models. Customer behavior can be captured by utilizing the discrete-choice models. However, selecting the right choice model is a very challenging problem. Significant portion of the research topic of this proposal is dedicated to selecting the right choice model and efficiently estimating the choice model parameters. Second portion of this research is on integrating estimated consumer behavior into revenue management algorithms. Using the approaches described in this proposal, the effective product and consumer matching will yield the right set of products to be offered to right set of customers, at the right time and at the right price. These are timely research topics due to both their promise for significant advances in a variety of applications, as well as our recent initial progress on these problems. We have played a significant role in that initial progress, hence we are in a very good position to lead the solution of the problems posed in this project.

Loading TOBB University of Economics and Technology collaborators
Loading TOBB University of Economics and Technology collaborators