Yönetim Bilişim Sistemleri Bölümü Koleksiyonu
Permanent URI for this collectionhttps://gcris.khas.edu.tr/handle/20.500.12469/68
Browse
Recent Submissions
Article Citation Count: 0Using Random Forest Tree Classification for Evaluating Vertical Cross-Sections in Epoxy Blocks to Get Unbiased Estimates for 3D Mineral Map(Gazi University, 2021) Camalan, Mahmut; Çavur, MahmutAreal mineral maps are constructed from the polished sections of particles that settle to the bottom of epoxy resin. However, heavy minerals can preferentially settle to the bottom, making the polished surface rich in heavy minerals. Therefore, polished sections will become biased estimates of the volumetric (3D) map. The study aims to test whether any vertical cross-section (any section along the settling direction of particles) can be an unbiased estimate of the 3D mineral map of a chromite ore sample. For the purpose of this study, 2D maps of the vertical cross-sections were acquired by using Random Forest classification coupled with image pre- and post-processing tools. Then, 3D mineral maps were converted from 2D maps without assuming stereological errors. The modal mineralogy and particle size distributions predicted from 3D maps were compared with the same features estimated from the particulate sample by XRD and dry sieving analyses, respectively. Any 2D map which yields a modal mineralogy and a size distribution similar to the true analyses was selected as an unbiased estimate of the true 3D map. The results suggest that any vertical cross-section is an unbiased estimate, unlike polished section at which heavier minerals settle preferentially.Book Part Citation Count: 12Comparison of post outage bus voltage magnitudes estimated by harmony search and differential evolution methods(2009) Ceylan, Oğuzhan; Dağ, Hasan; Dağ, HasanContingency studies are indispensable tools of both the power system planning and operational studies. Real time implementation of operational problems makes necessary the use of high speed computational methods while requiring reasonable accuracies. On the other hand, accuracy of the results and the speed of calculation depend on branch outage modeling as well as solution algorithm used. This paper presents a comparison of post outage bus voltage magnitudes calculated by two meta-heuristic approaches; namely differential evolution (DE) and harmony search (HS) methods. The methods are tested on IEEE 14, IEEE 30, IEEE 57, and IEEE 118 bus test systems and the results are compared both in terms of accuracy and calculation speed.Conference Object Citation Count: 0Comparison of Cost-Free Computational Tools for Teaching Physics(IEEE, 2010) Dağ, Hasan; Dağ, HasanIt is widely accepted that it is quite difficult to engage today's students, from high schools to university, both in educational activities in class and "teaching" them physics due to their prejudices about the complexity of physics. The difficulty in capturing students' attention in class for a long time also plays a role in less effective teaching during learning activities. Research shows that students learn little from traditional lectures. According to constructivist learning theories, visual aids and hands-on activities play a major role in learning physics. In addition to laboratory work there are many computational tools for teaching physics, which help teachers and students in constructing a conceptual framework. With this in mind, this paper compares freeware and open source computational tools for teaching physics.Article Citation Count: 5An inverse coefficient problem for a quasilinear parabolic equation with nonlocal boundary conditions(Springer International Publishing Ag, 2013) Kanca, Fatma; Bağlan, İremIn this paper the inverse problem of finding the time-dependent coefficient of heat capacity together with the nonlocal boundary conditions is considered. Under some natural regularity and consistency conditions on the input data, the existence, uniqueness and continuous dependence upon the data of the solution are shown. Some considerations on the numerical solution for this inverse problem are presented with an example.Article Citation Count: 0AN EFFECTIVE ROCOMMENDER MODEL FOR E-COMMERCE PLATFORMS(2017) Dağ, Hasan; Dağ, HasanSahte kullanıcı hesapları, veri tabalarındaki seyreklik problemlerinden dolayı özellikle yeteri kadar kullanıcı tarafından puanlanmamış ürünlerde tavsiye algoritmalarını kolaylıkla etkileyebilmektedirler. Genellikle bu kullanıcı hesapları kendi ürününün puanını artırmak isteyen ürün sahipleri olabildiği gibi herhangi bir ürünü veya şirketi karalamak isteyen kötü niyetli kişiler de olabilmektedir. Bu durum birçok şirketin veri tabanı yoğunluğunun %1 den daha az olduğu düşünülürse e-ticaret ortamlarına nasıl bir etki yarattığı tahmin edilebilir. Bu çalışmada, sahte hesapların e-ticaret ortamlarında oluşturdukları negatif etkilerin üstesinden gelebilmek için, kullanıcılar arasındaki ilişkiler analiz edilerek diğer kullanıcılar üzerinde etkisi olan ve gerçekten güvenilir olduğu düşünülen kullanıcılar bulunarak bir tavsiye modeli oluşturulmaktadır. Böylece, güvenilir kullanıcıların düşüncelerinden yola çıkılarak e-ticaret ortamlarında kullanıcılara tavsiyelerde bulunan Tavsiye Sistemlerinin (TS) kalitesini artıracak bir tavsiye sistemi oluşturulacaktırConference Object Citation Count: 1A comparative study of surrogate based learning methods in solving power flow problem(IEEE, 2020) Ceylan, Oğuzhan; Taşkın, Gülsen; Paudyal, SumitDue to increasing volume of measurements in smart grids, surrogate based learning approaches for modeling the power grids are becoming popular. This paper uses regression based models to find the unknown state variables on power systems. Generally, to determine these states, nonlinear systems of power flow equations are solved iteratively. This study considers that the power flow problem can be modeled as an data driven type of a model. Then, the state variables, i.e., voltage magnitudes and phase angles are obtained using machine learning based approaches, namely, Extreme Learning Machine (ELM), Gaussian Process Regression (GPR), and Support Vector Regression (SVR). Several simulations are performed on the IEEE 14 and 30-Bus test systems to validate surrogate based learning based models. Moreover, input data was modified with noise to simulate measurement errors. Numerical results showed that all three models can find state variables reasonably well even with measurement noise.Article Citation Count: 28Random CapsNet forest model for imbalanced malware type classification task(Elsevier, 2021) Dağ, Hasan; Ünal, Uğur; Dağ, HasanBehavior of malware varies depending the malware types, which affects the strategies of the system protection software. Many malware classification models, empowered by machine and/or deep learning, achieve superior accuracies for predicting malware types. Machine learning-based models need to do heavy feature engineering work, which affects the performance of the models greatly. On the other hand, deep learning-based models require less effort in feature engineering when compared to that of the machine learning-based models. However, traditional deep learning architectures components, such as max and average pooling, cause architecture to be more complex and the models to be more sensitive to data. The capsule network architectures, on the other hand, reduce the aforementioned complexities by eliminating the pooling components. Additionally, capsule network architectures based models are less sensitive to data, unlike the classical convolutional neural network architectures. This paper proposes an ensemble capsule network model based on the bootstrap aggregating technique. The proposed method is tested on two widely used, highly imbalanced datasets (Malimg and BIG2015), for which the-state-of-the-art results are well-known and can be used for comparison purposes. The proposed model achieves the highest F-Score, which is 0.9820, for the BIG2015 dataset and F-Score, which is 0.9661, for the Malimg dataset. Our model also reaches the-state-of-the-art, using 99.7% lower the number of trainable parameters than the best model in the literature.Book Part Citation Count: 1Open data in agriculture: Sustainable model development for hazelnut farms using semantics(Institute of Electrical and Electronics Engineers Inc., 2018) Aydın, Mehmet Nafiz; Ünal, Uğur; Aydın, Mehmet NafizTurkey accounts for 75% of the global hazelnut production and 70-75% of the exportation. Taking into account the socioeconomic importance of hazelnut, the stakeholders of hazelnut domain still have problems such as availability, meaningful, accuracy of the hazelnut related data. Providing data to stakeholders is crucial for sustainable agricultural activities. This data should be freely available to everyone to use and republish. With the aforementioned reasons "Open Data" is an efficient way in Turkish Agriculture.In this paper, we shall investigate the open data term and semantics in the context of hazelnut data management. In addition, a data processing model with regard to agricultural open data is proposed.Conference Object Citation Count: 26Feature Extraction Based on Deep Learning for Some Traditional Machine Learning Methods(Institute of Electrical and Electronics Engineers Inc., 2018) Dağ, Hasan; Yenidoğan, Işıl; Dağ, HasanDeep learning is a subfield of machine learning and deep neural architectures can extract high level features automatically without handcraft feature engineering unlike traditional machine learning algorithms. In this paper, we propose a method, which combines feature extraction layers of a convolutional neural network with traditional machine learning algorithms, such as, support vector machine, gradient boosting machines, and random forest. All of the proposed hybrid models and the above mentioned machine learning algorithms are trained on three different datasets: MNIST, Fashion-MNIST, and CIFAR-10. Results show that the proposed hybrid models are more successful than traditional models while they are being trained from raw pixel values. In this study, we empower traditional machine learning algorithms for classification using feature extraction ability of deep neural network architectures and we are inspired by transfer learning methodology to this.Conference Object Citation Count: 2A Country-specific Analysis on Internet Interconnection Ecosystems(IEEE, 2017) Aydın, Mehmet Nafiz; Aydın, Mehmet NafizWith the proliferating number of diverse participants and destinations to reach, the Internet construct has become more intricate to assay. Today, Internet Service Providers (ISPs) establish resilient networks from multiple providers and broaden the number of peering links - as financially as viable. However, the complex structure of the global Internet ecosystem and entwined roles of Internet players simply prevent us from conducting generalized models for grasping interconnections which could be applied globally regardless of the local surroundings. In this paper, the global inter-domain Internet topology is scrutinized by the help of interconnection characteristics within a country-specific stance. Our study on the Internet ecosystems helps us highlight the non-uniformity of interconnections by using both "real world" metrics and network science metrics. One of the significant findings that the analysis yields is that presence of well-established Internet Exchange Points (IXPs) in an interconnection ecosystem - besides the benefit of bolstering the peering fabric - increases the competitive nature of Internet transit market and boosts the inclination to multi-home for stub networks, thus increases the resilience of national Internet constructs.Conference Object Citation Count: 0Multi-Agent model of electricity networks - a perspective on distribution network charges(Institute of Electrical and Electronics Engineers Inc., 2019) Ceylan, Oğuzhan; Ceylan, OğuzhanIn the UK, DNOs are regulated by Ofgem to use two common distribution use of system charging methodologies: Common Distribution Charging Methodology and Extra-High Voltage Distribution Charging Methodology. To account for the changing landscape of the energy sector, Ofgem has recently published a consultation paper on changes to DUoS charging structure. This paper looks into the implications of distribution network charging in consumer-level adoption of low carbon technologies and vice-versa, using an agent-based model approach.Conference Object Citation Count: 19Grey Wolf optimizer for allocation and sizing of distributed renewable generation(Institute of Electrical and Electronics Engineers Inc., 2019) Ceylan, Oğuzhan; Ceylan, Oğuzhan; Özdemir, AydoğanIncreasing penetration of distributed energy resources (DERs) have brought operational and control philosophy changes in Smart Grids (SGs). Renewable energy based technologies are becoming more important due to their economic and environmental impacts. Distributed generations (DGs) in the form of small renewable energy resources such as solar photovoltaics (PVs) and Wind Turbines (WTs) are connected in radial distribution networks near to the loads. This paper presents optimal siting and sizing of distributed renewable energy resource to maintain voltage magnitude profiles. Bus voltage magnitude differences for each hour in a day of a distribution system are formulated as an objective function. Three consecutive days are taken into account representing the three seasons of a year. A new nature inspired algorithm Grey Wolf Optimizer (GWO) is used as a solution tool. The proposed formulation is applied to 33 bus and 69 bus radial distribution networks. MATLAB simulations are performed to validate the performance of the approach. Simulation results are discussed and compared with of the several available ones'.Conference Object Citation Count: 1Active and Reactive Power Load Profiling Using Dimensionality Reduction Techniques and Clustering(Institute of Electrical and Electronics Engineers Inc., 2019) Ceylan, Oğuzhan; Ceylan, Oğuzhan; Papadopoulos, Theofilos A.; Kazaki, Anastasia G.; Barzegkar-Ntovom, Georgios A.This paper proposes a methodology to characterize active and reactive power load profiles. Specifically, the approach makes use of fast Fourier Transform for conversion into frequency domain, principle component analysis to reduce the dimension and K-means++ to determine the representative load profiles. The data set consists of five-year measurements taken from the Democritus University of Thrace Campus. Test days were also classified as working and non-working. From the results it is observed that the proposed methodology determines representative load profiles effectively both regarding active and reactive power.Conference Object Citation Count: 18Land Use And Land Cover Classification Of Sentinel 2-A: St Petersburg Case Study(International Society for Photogrammetry and Remote Sensing, 2019) Çavur, Mahmut; Düzgün, Hafize Şebnem; Kemeç, Serkan; Demirkan, Doğa ÇağdaşLand use and land cover (LULC) maps in many areas have been used by companies, government offices, municipalities, and ministries. Accurate classification for LULC using remotely sensed data requires State of Art classification methods. The SNAP free software and ArcGIS Desktop were used for analysis and report. In this study, the optical Sentinel-2 images were used. In order to analyze the data, an object-oriented method was applied: Supported Vector Machines (SVM). An accuracy assessment is also applied to the classified results based on the ground truth points or known reference pixels. The overall classification accuracy of 83,64% with the kappa value of 0.802 was achieved using SVM. The study indicated that of SVM algorithms, the proposed framework on Sentinel-2 imagery results is satisfactory for LULC maps.Conference Object Citation Count: 11Optimal Allocation Of Multi-Type Distributed Generators For Minimization Of Power Losses In Distribution Systems(Institute of Electrical and Electronics Engineers Inc., 2019) Ceylan, Oğuzhan; Ceylan, Oğuzhan; Özdemir, AydoğanDistributed generation (DG), including photovoltaics (PVs), wind turbines (WTs) are becoming vital for Active Distribution Networks (ADN). Therefore, optimal sizing and allocation of these units can improve voltage profiles and reduce active power losses. This study concentrates on optimal allocation and sizing of two kinds of DG units (PVs and WTs) with 1 MW maximum size limits, due to the regulations in Turkey. The Whale optimization algorithm (WOA) and Grey wolf optimization algorithm (GWO) are used as optimization tools to minimize active power losses of 33 and 69 Bus Test Systems. The performance analysis of the methods are performed through simulations and the numerical results are compared in terms of optimal OF values and convergence characteristics.Conference Object Citation Count: 1Metadata Action Network Model for Cloud Based Development Environment(Springer, 2020) Aydın, Mehmet Nafiz; Perdahçı, Ziya Nazım; Şafak, İlker; van Hillegersberg, JosCloud-based software development solutions (entitled as Platform-as-a-Service, Low-Code platforms) have been promoted as a game changing paradigm backed by model-driven architecture and supported by various cloud-based services. With the engagement of a sheer number of platform users (experienced, novel, or citizen developers) these platforms generate invaluable data and that can be considered as user metadata actions. As cloud-based development solutions provide novice users with a new development experience (performing data actions that altogether leads to a successful software app), users often times face with uncertainty about development performance; how good or complete is app development? Thus, the issue addressed in this research is how to measure user performance by using digital trace data generated on the cloud platform from a Network Science perspective. This research proposes a novel approach to leveraging digital trace data on Platform-as-a-Service (PaaS) from a Network Science perspective. The proposed approach considers the importance of digital trace data as metadata actions on PaaS and introduces a network model (so-called Metadata Action Network), which is claimed to be the result of reconstruction of events of developer’s actions. We show suitability of the proposed approach to better understanding of real-world digital trace data on PaaS solution and elaborate basic performance analytics on a PaaS solution with research and practical implications.Article Citation Count: 8A Sustainable Multi-Layered Open Data Processing Model For Agriculture: IoT Based Case Study Using Semantic Web For Hazelnut Fields(ASTES Publishers, 2020) Aydın, Mehmet Nafiz; Aydın, Mehmet NafizIn recent years, several projects which are supported by information and communications technologies (ICT) have been developed in the agricultural domain to promote more precise agricultural activities. These projects account for different kinds of key ICT terms such as internet of things (IoT), wireless sensors networks (WSN), cloud computing (CC). These projects are used for different agricultural products; and it is a well-known fact that they can be essential to perform precise agricultural activities for the relevant agricultural products. The implementation of these projects successfully depends on the extent to which various stakeholders provide support by leveraging relevant data, gathered from heterogenous data sources. Agriculture domain has a great number of stakeholders. These stakeholders need sophisticated data and appropriate intelligence to get benefits in order to perform precise agricultural activities. Authors agreed with scholars that "Open Data" idea, which means accessing data published on the web and available in a machine-readable format is an appropriate way to get benefits for precise agriculture by relevant stakeholders. In this paper, authors shall investigate the open data term in an agricultural context, create an open data processing model, and develop an IoT-based solution to gather environmental data from agricultural fields. Authors also show viability of the proposed model by developing an ICT-based solution. Considering the socioeconomic importance of hazelnut for Turkey, the stakeholders of hazelnut domain still have problems such as availability, meaningful, accuracy of the hazelnut related data. Therefore, authors shall focus on hazelnut within the scope of this paper.Conference Object Citation Count: 0Optimization of Graph Affinity Matrix with Heuristic Methods in Dimensionality Reduction of Hypespectral Images(IEEE, 2019) Ceylan, Oğuzhan; Taşkın, GülşenHyperspectral images include hundreds of spectral bands, adjacent ones of which are often highly correlated and noisy, leading to a decrease in classification performance as well as a high increase in computational time. Dimensionality reduction techniques, especially the nonlinear ones, are very effective tools to solve these issues. Locality preserving projection (LPP) is one of those graph based methods providing a better representation of the high dimensional data in the low-dimensional space compared to linear methods. However, its performance heavily depends on the parameters of the affinity matrix, that are k-nearest neighbor and heat kernel parameters. Using simple methods like grid-search, optimization of these parameters becomes very computationally demanding process especially when considering a generalized heat kernel, including an exclusive parameter per feature in the high dimensional space. The aim of this paper is to show the effectiveness of the heuristic methods, including harmony search (HS) and particle swarm optimization (PSO), in graph affinity optimization constructed with a generalized heat kernel. The preliminary results obtained with the experiments on the hyperspectral images showed that HS performs better than PSO, and the heat kernel with multiple parameters achieves better performance than the heat kernel with a single parameter.Conference Object Citation Count: 0GRAPH OPTIMIZED LOCALITY PRESERVING PROJECTION VIA HEURISTIC OPTIMIZATION ALGORITHMS(IEEE, 2019) Ceylan, Oğuzhan; Taşkın, GülşenDimensionality reduction has been an active research topic in hyperspectral image analysis due to complexity and non-linearity of the hundreds of the spectral bands. Locality preserving projection (LPP) is a linear extension of the manifold learning and has been very effective in dimensionality reduction compared to linear methods. However, its performance heavily depends on construction of the graph affinity matrix, which has two parameters need to be optimized: k-nearest neighbor parameter and heat kernel parameter. These two parameters might be optimally chosen simply based on a grid search when using only one representative kernel parameter for all the features, but this solution is not feasible when considering a generalized heat kernel in construction the affinity matrix. In this paper, we propose to use heuristic methods, including harmony search (HS) and particle swarm optimization (PSO), in exploring the effects of the heat kernel parameters on embedding quality in terms of classification accuracy. The preliminary results obtained with the experiments on the hyperspectral images showed that HS performs better than PSO, and the heat kernel with multiple parameters achieves better performance than the isotropic kernel with single parameter.Conference Object Citation Count: 0Impacts of Load and Generation Volatilities on the Voltage Profiles Improved by Distributed Energy Resources(Institute of Electrical and Electronics Engineers Inc., 2020) Ceylan, Oğuzhan; Ceylan, Oğuzhan; Özdemir, AydoğanWeather-dependent distributed renewable energy sources such as photovoltaics (PVs) and wind turbines (WT) are increasingly being connected to distribution networks (DNs). Increased penetration of these intermittent sources brought the necessity of using energy storage systems (ESSs) to achieve the intended benefits. This study presents an optimization process to determine optimal numbers, sizes, locations and distributed energy resources (DERs) as well as to determine the optimal operating strategy of ESSs in a distribution network. The objective is to improve the voltage profile and to minimize the installation costs. The proposed multi-objective formulation problem is solved by using ant lion multi-objective optimization algorithm. At the second part of the study, optimal values are tested with monthly extreme distributions and the impacts of load and distributed generation volatilies on the voltage profiles which were determined by Pareto-optimal solution candidates are analysed. Simulations were performed on 33 bus radial distribution system using Matlab. Finally the benefits obtained by the optimal solutions with less risk are compared.
