Search

Conference Name: 2022 30th Signal Processing and Communications Applications Conference (SIU) Date of Conference: 15-18 May 2022 Bir çok büyük veri işleme uygulaması “PageRank”, “İşbirliğine Dayalı Filtreleme” ve “Betweenness Centrality” gibi düzensiz ve yinelemeli çizge algoritmalarından yararlanmaktadır. Çizge yapılarında her bir düğüm bir kişiye ya da nesneye karşılık gelirken, her bir ayrıt da bir kişi ya da nesne çifti arasındaki ilişkiye karşılık gelmektedir. Böylesi büyük verileri işleyen algoritmaları yürütecek genel amaçlı işlemciler yetersiz kalabilmektedir. Özellikle, güç kısıtları sebebiyle işlemci alt parçalarından sadece bir kısmı aynı anda aktif olarak kullanılabilmektedir. Bu da işlemcinin etkin olarak kullanılmasına engel olarak aktif olmayan “karanlık silikon”lar ortaya çıkarmaktadır. Bu bildirinin amacı da büyük verili ve düzensiz çizge uygulamalarını hızlı, verimli ve kolay programlanabilir biçimde çalıştıracak bir işlemci mimarisi tasarlamaktır. Literatürde çizge uygulamaları için daha önce sunulmuş çalışmalar çoğunlukla sadece yazılım ya da hızlandırıcı seviyesindedir. İşlemci seviyesi tasarımlar ise donanımsal maliyet, talep ettikleri mimari destek ve komut seti değişiklikleri açılarından bu bildiri kapsamında tanıtılacak işlemciden farklıdır. Many widely used big data applications make use of irregular and iterative graph processing algorithms like PageRank, collaborative filtering and betweenness centrality. In a graph structure, each vertex corresponds to a person or object, while each edge corresponds to the relationship between a pair of persons or objects. In executing algorithms that process such large datasets, general-purpose processors may be inefficient. In particular, due to power constraints, only a portion of the processor can be used simultaneously. This, in turn, prevents the efficient use of the processor, which leads to inactive areas termed as "dark silicone". The aim of this paper is to design a processor architecture that will run large datasets on irregular graphs quickly, efficiently and in an easily programmable fashion. Most of the studies previously presented for graph applications are only at the software or accelerator level, and the processor-level designs are different from the processor to be introduced for this paper in terms of hardware cost, required architectural support, and instruction set modifications.

Conference Name: 2022 30th Signal Processing and Communications Applications Conference (SIU) Date of Conference: 15-18 May 2022 The performance of object detection algorithms running on images taken from Unmanned Aerial Vehicles (UAVs) remains limited when compared to the object detection algorithms running on ground taken images. Due to its various features, YOLO based models, as a part of one-stage object detectors, are preferred in many UAV based applications. In this paper, we are proposing novel architectural improvements to the YO-LOv5 architecture. Our improvements include: (i) increasing the number of detection layers and (ii) use of transformers in the model. In order to train and test the performance of our proposed model, we used VisDrone and SkyData datasets in our paper. Our test results suggest that our proposed solutions can improve the detection accuracy. İnsansız Hava Araçları (İHA) üzerindeki kameralardan alınan görüntülerde, nesne tanıma algoritmalarının başarımı, yerden çekilen görüntülerdeki performanslara göre daha sınırlı kalabilmektedir. Birçok İHA uygulamasında, tekgeçişli olmasından ve iki-geçişli yöntemlere göre daha hızlı olmasından dolayı, önceden eğitilmiş modelleri ile beraber gelen YOLOv5 algoritması kullanılmaya başlanmıştır. Ancak YOLOv5 algoritmasının İHA tabanlı imgeler üzerindeki performansı, yerden çekilen resimler üzerindeki performansına oranla düşük kalmaktadır. O nedenle bu çalışmada, YOLOv5 algoritmasının performansını arttırmak adına mimari çözümler önerilmiştir. Bu çözümler: (i) farklı sayıda tahmin yapan katman kullanmak, ve (ii) dönüştürücü (transformer) katmanı kullanmak olarak sıralanabilir. Sunulan yöntem, en güncel olan VisDrone ve Sky- DataV1 veri kümeleri üzerinde, orijinal YOLOv5 algoritması ile karşılaştırıldığında, performans artışı sağlamıştır.

Date of Conference: 16-19 May 2016 Conference Name: IEEE 24th Signal Processing and Communications Applications Conference, SIU 2016 In this paper, we study multi-class classification of tweets, where we introduce highly efficient dimensionality reduction techniques suitable for online processing of high dimensional feature vectors generated from freely-worded text. As for the real life case study, we work on tweets in the Turkish language, however, our methods are generic and can be used for other languages as clearly explained in the paper. Since we work on a real life application and the tweets are freely worded, we introduce text correction, normalization and root finding algorithms. Although text processing and classification are highly important due to many applications such as emotion recognition, advertisement selection, etc., online classification and regression algorithms over text are limited due to need for high dimensional vectors to represent natural text inputs. We overcome such limitations by showing that randomized projections and piecewise linear models can be efficiently leveraged to significantly reduce the computational cost for feature vector extraction from the tweets. Hence, we can perform multi-class tweet classification and regression in real time. We demonstrate our results over tweets collected from a real life case study where the tweets are freely-worded, e.g., with emoticons, shortened words, special characters, etc., and are unstructured. We implement several well-known machine learning algorithms as well as novel regression methods and demonstrate that we can significantly reduce the computational complexity with insignificant change in the classification and regression performance.

National Conference on Electrical, Electronics and Biomedical Engineering (ELECO) (2016 : Bursa, Turkey), Nowadays, unmanned aerial vehicles became an important research area. Search and rescue, transportation, fire monitoring, commercial and journalism, military and defence, crop spraying and scientific research are some of the areas which UAVs can be used [1,2]. Because of their agile manoeuvrability, quadcopters are often used for controller designs. In this paper a sliding mode controller is developed for altitude and attitude control, and the obtained simulation results are discussed. After the design is made, coefficients are tuned by iteration method. Also a drawback called chattering effect is explained and reducted to achieve better control. The rest of the paper provides more detailed information about abovementioned topics.

Conference Name: 2022 30th Signal Processing and Communications Applications Conference (SIU) Date of Conference: 15-18 May 2022 Functional magnetic resonance imaging (fMRI) enables recording the brain’s neural activity spatiotemporally and is the center of much cutting-edge psychology and neuroscience research. Many methods are proposed to process the 4-dimensional data the fMRI scans provide. The most common approach for classification tasks is to analyze functional connectivity, where brain volume is parcelled to regions, and the correlation between their time series is calculated. Such an approach is very suitable for graphical neural networks, a popular deep learning method for graphical data analysis. A graph is constructed by formulating the parcelled brain regions as the graph nodes, while their features and edges are constructed from the correlations. However, in many studies, the correlations are calculated from simple methods that do not take account of the lagged relations between the node time-series. This paper addresses this issue by proposing a new graphical neural network layer. This layer accounts for lagged relationships between the nodes and learns reacher features rather than simple zero-lag correlations. We show that our graphical layer can be used in front of a known graphical model and increase its performance for two different downstream tasks in a large fMRI dataset. Fonksiyonel manyetik rezonans görüntüleme (fMRG), beyindeki sinirsel etkinliği zamansal ve uzamsal olarak kaydedebilen bir görüntüleme tekniğidir ve yenilikçi psikoloji ve sinirbilimi araştırmalarının merkezindedir. fMRG taramalarının 4 boyutlu verisini işleyebilmek için çeşitli metotlar önerilmiştir. Sınıflandırma çalışmalarında en yaygın olarak kullanılan teknik, beynin bölgelere ayrılması ve bu bölgelerin zaman serileri arasında korelasyon hesaplanmasıyla bulunan fonksiyonel bağlılık ölçütüdür. Söz konusu yaklaşım, grafiksel verilerin derin öğrenme ile işlenmesinde popüler bir teknik olan grafiksel sinir ağlarında kullanmak için uygundur. Grafiksel sinir ağlarında bölünmüş beyin bölgeleri düğümleri oluştururken düğümler arasındaki bağlantılar ve düğümlerin özellik vektörleri korelasyon hesabına dayanır. Çoğu çalışmada bu korelasyon hesabı yapılırken dü- ğümlerin zaman serileri arasındaki gecikmeli ilişkiler göz ardı edilmektedir. Bu makalede önerilen yeni sinirsel ağ katmanıyla gecikmeli ilişkilerin etkisinin incelenmesi hedeflenmiştir. Bu katman düğümler arasında gecikmeli ilişki hesabı yaparak basit, sıfır gecikmeli korelasyona göre daha zengin özellik vektörleri oluşturulmasını sağlar. Bu makaleyle, önerdiğimiz grafiksel katmanın bilinen başka bir grafiksel modelin önüne eklenmesi sonucu performans artımı sağlanabileceğini 2 çalışmayla gösteriyoruz.

Conference Name: 2022 30th Signal Processing and Communications Applications Conference (SIU) Date of Conference: 15-18 May 2022 The Batch-Constrained Q-learning algorithm is shown to overcome the extrapolation error and enable deep reinforcement learning agents to learn from a previously collected fixed batch of transitions. However, due to conditional Variational Autoencoders (VAE) used in the data generation module, the BCQ algorithm optimizes a lower variational bound and hence, it is not generalizable to environments with large state and action spaces. In this paper, we show that the performance of the BCQ algorithm can be further improved with the employment of one of the recent advances in deep learning, Generative Adversarial Networks. Our extensive set of experiments shows that the introduced approach significantly improves BCQ in all of the control tasks tested. Moreover, the introduced approach demonstrates robust generalizability to environments with large state and action spaces in the OpenAI Gym control suite. Toplu-Kısıtlı Q-öğrenme (TKQ) algoritmasının, ekstrapolasyon hatasının üstesinden geldiği ve derin pekiştirmeli öğrenme ajanlarının önceden toplanmış sabit bir deneyim kümesinden öğrenebildiği gösterilmiştir. Ancak, veri oluşturma modülünde kullanılan Koşullu Değişken Özkodlayıcılar (KDÖ) sebebiyle TKQ algoritması daha düşük bir değişken sınırını optimize etmektedir ve bu nedenle, büyük durum ve eylem uzaylarına sahip ortamlara genelleştirilememektedir. Bu bildiride, derin öğrenmedeki son gelişmelerden biri olan Üretken Çekişmeli Ağlar’ın (ÜÇA) kullanılmasıyla TKQ algoritmasının performansının daha da geliştirilebileceği gösterilmektedir. Kapsamlı deneyler, tanıtılan yaklaşımın test edilen her bir kontrol görevinde TKQ’yu önemli ölçüde geliştirdiğini göstermektedir. Ayrıca tanıtılan yaklaşım, OpenAI Gym kontrol setindeki geniş durum ve eylem boşluklarına sahip ortamlara hızlı bir genelleştirilebilirlik sergilemektedir.

Bu çalışmada LoRa tabanlı modülasyonların toplamsal beyaz Gauss ve Rayleigh kanallardaki bit hata oranı başarımları incelenmiştir. Nominal LoRa, ICS-LoRa ve SSKLoRa modülasyonlarının mesaj çözüm kuralları, demodülatör karmaşıklıkları ve spektral verimlilikleri karşılaştırılmıştır. Bit hata oranı sayısal sonuçları verilmiştir., In this paper, the bit error rate (BER) performances of the LoRa-based modulations are investigated for the additive white Gaussian noise channel and Rayleigh channel. Detection rules, demodulator complexity and spectral efficiency of the nominal LoRa, the interleaved chirp spreading LoRa (ICS-LoRa), and the slope-shift-keying LoRa (SSK-LoRa) modulations are compared. BER numerical results are given.

The microwave imaging systems have the potential to present a cost effective and less hazardous alternative to conventional medical imaging techniques. In this paper, a Contrast Source Inversion method based microwave imaging scheme is proposed and tested for the detection of pulmonary edema and hemorrhage. To this end, a realistic human torso phantom is used, and the electromagnetic parameters of the human tissues is determined via Cole-Cole model. The scattered field is simulated via Method of Moments at the operating frequency of 350 MHz, and a 50 dB white Gaussian noise is added to model a realistic measurement setup. The numerical tests performed with the proposed technique suggest that the method can be used to locate the pulmonary edema and hemorrhage, and it is capable of distinguishing these two medical conditions successfully. © 2022 IEEE. 2-s2.0-85138721822

Conference Name: 2021 29th Signal Processing and Communications Applications Conference (SIU) Date of Conference: 9-11 June 2021 Bu bildiride, i) yapay zeka ile öğrenme amaçlı geliştirilen DeepRTS oyunu için makro aksiyonların kullanıldığı bir çevre oluşturulmuş ve ii) belirli pekiştirmeli öğrenme algoritmaları, üzerlerinde gerekli değişiklikler yapılarak bu çevre üzerinde eğitimleri sağlanmış ve ilgili performans analizleri yapılmı ştır. Gerçek hayat planlamalarıyla paralellik çizme ve görece düşük donanımlarla bir gerçek zamanlı strateji oyununu oynama amacıyla DeepRTS oyunu üzerinde değişikliklere gidilmiştir. ˙Ilk olarak, bir makro aksiyon seti hazırlanmış ve ajanların yalnızca bu set içerisinden aksiyon alabilmesi sağlanmıştır. ˙Ikinci olarak, gerçek hayat planlamalarına paralel olarak sistem herhangi bir anda komut alabilecek bütün birimler için aksiyon alınabilecek duruma getirilmiştir. Bu durum makro aksiyonların farklı zaman adımları sürmesi ile birlikte ele alındığında, herhangi bir anda birden fazla aksiyonun başlayıp, birden fazla aksiyonun tamamlanmasına olanak sağladığı için klasik pekiştirmeli öğrenme probleminden bir miktar ayrılmıştır. Ayrıca, literatürde bilinen kredi atama problemine farklı bir boyut ekleyerek daha karmaşık hale getirmektedir. Ajanların eğitiminde kullanılma amacı ile ofansif, defansif ve rastgele kural tabanlı ajanlar oluşturulmuş ve pekiştirmeli öğrenme tabanlı ajanların eğitimleri sırasında dönüşümlü şekilde düşman ajan olarak kullanılmıştır. Eğitimi tamamlanan ajanların birbirlerine karşı oyuncu-1 ve oyuncu-2 olarak performansı raporlanmıştır. In this paper, i) we build a framework using macro actions on top of DeepRTS, which is specifically developed for learning purposes, and ii) we train a number of deep reinforcement learning based agents and then we conduct a detailed performance analysis on the performance of these agents. We make changes on the publicly available version of DeepRTS to make it parallel to real life planning problems and make it trainable with a decent hardware. Firstly, we create a set of macro actions based on human heuristics and we train our agent using only these actions. Secondly, similar to the real life planning problem, we change the problem such that we simultaneously take actions for all available unit at a time step. When we consider this situation together with duration of macro actions being different, multiple actions may start and multiple actions ends at any time step. Therefore, the problem draw apart from classical reinforcement learning problem in a certain amount. In addition, the credit assignment problem becomes more sophisticated since each action lasts a duration and at any time multiple actions are conducted. We create three rule based agents to use them as enemy players in training of our agents. We report the performance of the agents as player-1 and player-2 against each others.

29th IEEE Conference on Signal Processing and Communications Applications (SIU) -- JUN 09-11, 2021 -- ELECTR NETWORK, Today, thanks to the developing technology, changes are experienced in many areas and these changes sometimes bring some problems, issues that need to be solved or improved. In the field of digital image and video processing, the viewer experience has been greatly improved with the development of image and video coding methods and technologies in recent years. However, due to these developments, the importance of video and image quality evaluation techniques has increased. In the field of imaging technology, although the human visual system and human judgment experiments will give us the most accurate results, this rapidly growing field has made human judgment experiments unsuitable with the increasing content. At this point, effective and automatic quality assessment metrics are needed to evaluate and optimize advanced compression technologies in terms of visual perception quality. In this paper, our aim is to compare the performances of some of the new generation image and video quality assessment metrics., IEEE,IEEE Turkey Sect

15th Turkish National Software Engineering Symposium (UYMS) -- NOV 17-19, 2021 -- ELECTR NETWORK, The specific requirements of developing Internet of Things (IoT) software for public transportation systems should be considered in addition to the problems always observed in IoT system development. After defining the problems encountered in the development of IoT-based public transport systems, the current model-driven engineering (MDE) approaches for IoT development are examined and their applicability for public transport systems is questioned in this paper. Several points of view are given about what new MDE approaches specific to this field should include in order to eliminate the identified deficiencies. Moreover, it is indicated that new MDE approaches should be designed including round-trip engineering to provide high mobility, use of special standards, in-vehicle communication networks and support for different system-on-chips in these public transportation systems., Ege Univ,Izmir Inst Technol,Gazi Univ,IEEE Turkey Sect,Huawei,Kuveyt Turk Arge Merkezi,Logo,Dalya Turizm Kongre Organizasyon

In this paper the application of a multigrid based sparse recovery method for multidimensional harmonic retrieval is investigated. Multidimensional modal estimation problem is formulated as a simultaneous sparse recovery problem for each dimension and then the signal parameters are estimated using a multigrid dictionary refinement scheme. Performance of the proposed method is compared against 2-D ESPRIT and Cramer-Rao lower bounds in numerical simulations., Bu çalışmada çok ızgaralı seyrek geriçatım yönteminin çok boyutlu sinyallerin kestirimine uygulanması incelenmiştir. Çok boyutlu sinyallerin bileşenlerinin kestirilmesi problemi herbir boyutu için aynı anda seyrek geriçatım problemine dönüştürüldükten sonra sinyalin o boyuttaki bileşen kestirimi çok ızgaralı sözlük matrisi güncellemeyle iyileştirilmektedir. Önerilen yöntemin başarımı, 2-B ESPRIT yöntemi ve Cramer-Rao alt sınırı ile sayısal benzetimlerle karşılaştırılmıştır.

Date of Conference: 24-26 April 2019 Conference name: 27th Signal Processing and Communications Applications Conference (SIU), 2019 Bu bildiride, gözlemlenen verideki anomalileri, gözetimsiz bir çerçevede, iki aşamalı yöntemle bulmak için anomali tespit algoritması tanıtılmıştır. İlk aşamada, ardışık olarak gözlemlenen verinin yoğunluğu çekirdek temelli özgün bir yöntemle tahmin edilmektedir. Bu amaçla, gözlem alanı bölünmekte ve her bölgede parametrik olmayan Çekirdek Yoğunluk Tahmincisi (ÇYT) veri dağılımına dair hiçbir varsayımda bulunulmadan kullanılmaktadır. Sonra, yoğunluk tahmini eşik değeriyle karşılaştırılarak verinin anomali olup olmadığına karar verilmektedir. Ayrıca, çekirdek temelli yöntemlerdeki bant genişliği seçimi problemi de verimli bir şekilde çözülmektedir. Bu amaçla, her bir bölgeye çekirdek bant genişliği seti atanmakta ve her tahmincinin ait olduğu bölgeye göre en iyi bant genişliği seçeneğine zamanla ulaşması sağlanmaktadır. Sayısal örneklerde, tanıtılan algoritmanın literatürde sıklıkla kullanılan anomali tespit metodlarına göre yüksek performans artışı elde ettiği gösterilmektedir. In this paper, we introduce an online anomaly detection algorithm to detect the anomalies in the observed data with two step approach in an unsupervised framework. In the first step, we estimate the density of the sequentially observed data with a novel kernel based approach. To this end, we partition the observation space and use nonparametric Kernel Density Estimator (KDE) in each region on a partition such that we do not assume any underlying distribution for the data. Then, we compare the estimated density of the data with a threshold to decide whether it is anomalous. We also solve the bandwidth selection problem in kernel based approaches in an efficient way. For this, we assign a set of kernel bandwidth values to each region, and make each estimator to converge to the best bandwidth choice for the corresponding subspaces in time. In our experiments, we show that our algorithm significantly outperforms the anomaly detection algorithms, which are highly used in the literature.

Bu makalede bir DA-DA düşürücü dönüştürücü sistemindeki oransal-integral-türevsel (PID) denetleyici parametrelerinin optimum değerlerini belirlemek için balina optimizasyon algoritmasını (WOA) kullanan yeni bir tasarım yöntemi sunulmaktadır. Önerilen algoritmanın basitliği optimum PID denetleyici parametrelerinin etkin bir şekilde hızlı ve yüksek kalitede ayarlanmasını sağlar. Önerilen WOA-tabanlı PID denetleyicinin performansı bir zaman domeni performans ölçütü kullanılarak doğrulanmıştır. Benzetim sonuçlarından, önerilen yöntemin genetik algoritmaya (GA) kıyasla DA-DA düşürücü dönüştürücünün geçici hal cevabını iyileştirmede daha etkin olduğu bulunmuştur., This paper presents a novel design method to determine optimal proportional-integral-derivative (PID) controller parameters of a DC-DC buck converter using the whale optimization algorithm (WOA). The simplicity of the proposed algorithm provides fast and high quality tuning of optimum PID controller parameters effectively. The performance of the proposed WOA-based PID controller is validated using a time domain performance index. From simulation results, the proposed method compared to the genetic algorithm was found more efficient in improving the transient response of Buck converter.

Bu makalede salp sürüsü algoritması (SSA) adında yeni bir yapay zekaya dayalı optimizasyon metodu otomatik gerilim regülatörü (AVR) sisteminin en uygun oransal, integral, türevsel (PID) kontrolör parametrelerinin belirlemesi amacıyla kullanılmıştır. Algoritmanın basitliği, optimal PID kontrolör parametrelerinin yüksek kaliteli ayarını sağlar. Kontrolör parametrelerinin optimize edilmesi için zaman ağırlıklı karesel hatanın integrali (ITSE) amaç fonksiyonu olarak seçildi. Geçici hal cevap analizi, SSA metodunun Ziegler-Nichols (ZN) geleneksel ayarlama yönteminden ve yapay arı kolonisi (ABC) algoritmasından daha iyi bir ayarlama kabiliyetine sahip olduğunu ve bir AVR sisteminin basamak cevabını iyileştirmede daha verimli olduğunu ortaya koymuştur., — In this paper, a novel artificial intelligence based optimization method, called salp swarm algorithm (SSA) is applied to determine the optimal proportional, integral, derivative (PID) controller parameters of the automatic voltage regulator (AVR) system. The simplicity of algorithm provides high quality tuning of optimal PID controller parameters. The integral of time weighted squared error (ITSE) as objective function is selected to optimize the controller parameters. The transient response analysis reveals that the SSA method has a better tuning ability and is more efficient in improving the step response of an AVR system than Ziegler-Nichols (ZN) conventional tuning method and artificial bee colony (ABC) algorithm.

İlişkisel veri analizinde temel bir problem olan topluluk seziminde, ilişkisel veri için model seçimi hala açık bir problemdir. Bu bildiride, Bayesçi atama modelinin (BAM) genel atama çerçevesi içinde, karmaşık üyelik rastlantısal öbek modellerinin (KÜRÖB) model boyutunun tahmin edilmesi önerilmektedir. İlişkisel verilerin atama modelindeki Poisson sayıları olarak nasıl temsil edildiği açıklanmaktadır. Deney sonuçları sentetik ve gerçek veri kümeleri üzerinde gösterilmektedir. Genelleyici atama yaklaşımı, sadece model boyutunu değil, aynı zamanda en uygun ayrıştırmayı seçebilen genelleştirilmiş bir model seçimi çözümü vaat etmektedir Abstract—As a fundamental problem in relational data analysis, model selection for relational data factorization is still an open problem. In our work, we propose to estimate model order for mixed membership blockmodels (MMSB) within the generic allocation framework of Bayesian allocation model (BAM). We describe how relational data is represented as Poisson counts of the allocation model, and demonstrate our results both on synthetic and real-world data sets. We believe that the generic allocation perspective promises a generalized model selection solution where we do not only select the model order, but also choose the most appropriate factorization. IEEE Turkey Sect; Turkcell; Turkhavacilik Uzaysanayii; Turitak Bilgem; Gebze Teknik Univ; SAP, Detaysoft; NETAS; Havelsan WOS:000518994300248 2-s2.0-85079770210 Conference Proceedings Citation Index- Science Proceedings Paper Nisan 2019 YÖK - 2018-19

Hyperspectral 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.

Çalışmamızın amacı Türkçe için paragraf-cümle düzeyinde anlamsal söylem analizi ve paragraf-cümle ve cümle-cümle düzeyinde metinsel benzerlik ölçümlemesi için bir veri kümesi hazırlamaktır. The aim of paper is generate a data set for analysing of Turkish semenatic and sentence similarity. Publisher's Version

Conference name: 26th Signal Processing and Communications Applications Conference (SIU) 2018 Date of Conference: 2-5 May 2018 Görüntü ters evrişimi görüntülemede sık karşılaşılan ters problemlerden biridir. Doğal görüntüler bir dönüşüm uzayında seyrek olarak ifade edilebildikleri için seyreklik önsellerinin bu problemleri etkili bir şekilde düzenlileştirebildikleri ve yüksek kaliteli yeniden oluşturmayı mümkün kıldıkları bilinmektedir. Bu çalışmada, görüntü ters evrişimi için dönüşüm öğrenmeye dayalı adaptif bir seyrek görüntü geri kazanım yaklaşımı geliştirilmektedir. Bu yaklaşımda adaptif olarak yama tabanlı bir seyrekleştirici dönüşüm öğrenilir ve eş zamanlı olarak gürültülü, bulanık ölçümlerden görüntü geri kazanılır. Ortaya çıkan optimizasyon problemi kapalı formda ifade edilebilen ve hızlı bir şekilde hesaplanabilen adımlar içeren almaşık enküçültme (alternating minimization) algoritması kullanılarak çözülmektedir. Geliştirilen algoritmanın performansı, farklı optik bulanıklıkları ve gürültü seviyeleri için incelenmektedir. Sonuçlar, geliştirilen yöntemin toplam değişim (total variation) temelli yaklaşımla karşılaştırıldığında, görüntü geriçatım kalitesini arttırdığını, aynı zamanda da hızlı olduğunu göstermektedir. Image deconvolution is one of the most frequently encountered inverse problems in imaging. Since natural images can be modeled sparsely in some transform domain, sparsity priors have been shown to effectively regularize these problems and enable high-quality reconstructions. In this paper, we develop a data-adaptive sparse image reconstruction approach for image deconvolution based on transform learning. Our framework adaptively learns a patch-based sparsifying transform and simultaneously reconstructs the image from its noisy blurred measurement. This is achieved by solving the resulting optimization problem using an alternating minimization algorithm which has closed-form and efficient update steps. The performance of the developed algorithm is illustrated for an application in optical imaging by considering different optical blurs and noise levels. The results demonstrate that the developed method not only improves the reconstruction quality compared to the totalvariation based approach, but also is fast.

International Congress on Big Data, Deep Learning and Fighting Cyber Terrorism (IBIGDELFT) -- DEC 03-04, 2018 -- Turkish IT Author, Ankara, TURKEY WOS: 000459239400016 In this paper, hardware efficient diffusion layers are generated. Non-involutory and new 8x8 MDS matrices over GF(2(8)) with the best-known XOR counts are constructed by using a new matrix form, which is developed by circulant matrix form and the idea used in developing Generalised Hadamard (GHadamard) matrix form. Then, it is shown that the proposed matrix form is a subset of Toeplitz matrix form. Gazi Univ, Minist Transportat & Infrastrucuture Turkey, Havelsan, Aselsan, BiSoft, Oracle, Proda, Netas, RStudio, Cisco, IEEE Turkey Sect, Informat & Commun Technologies Author

Date of Conference: 2-5 May 2018 In this paper, we study online nonlinear data regression and propose a highly efficient long short term memory (LSTM) network based architecture. Here, we also introduce on-line training algorithms to learn the parameters of the introduced architecture. We first propose an LSTM based architecture for data regression. To diminish the complexity of this architecture, we use an energy efficient operator (ef-operator) instead of the multiplication operation. We then factorize the matrices of the LSTM network to reduce the total number of parameters to be learned. In order to train the parameters of this structure, we introduce online learning methods based on the exponentiated gradient (EG) and stochastic gradient descent (SGD) algorithms. Experimental results demonstrate considerable performance and efficiency improvements provided by the introduced architecture.

Medical Technologies National Congress (TIPTEKNO) -- NOV 08-10, 2018 -- Magusa, CYPRUS WOS: 000467637600004 In this paper, a post-operative recurrence estimation tool called Sorbellini's nomogram for the kidney cancer patients showing no metastates is introduced and a novel application for mobile devices based on this model is developed for the physician's follow up procedures. The TNM stage, tumor size, nuclear (Fuhrman) grade, the existance of necrosis and vascular invasion are employed as the input parameters for this software to predict the recurrence probability in mentioned patients. Finaly, the performance analyses are carried out to verify the reliability of the application. Biyomedikal Klinik Muhendisligi Dernegi, Izmir Katip Celebi Univ, Biyomedikal Muhendisligi Bolumu

3rd International Conference on Computer Science and Engineering (UBMK) -- SEP 20-23, 2018 -- Sarajevo, BOSNIA & HERCEG, WOS: 000459847400034, Bone drilling operation is widely used method in surgery. This operation must be done with hight care. In the case of failur some demage might be happened. In this paper a new technich is presented. By this method 1,5 mm precision is optained., BMBB, Istanbul Teknik Univ, Gazi Univ, ATILIM Univ, Int Univ Sarajevo, Kocaeli Univ, TURKiYE BiLiSiM VAKFI

Date of Conference: 2-5 May 2018 Compressive spectral imaging is a rapidly growing area yielding higher performance novel spectral imagers than conventional ones. Inspired by compressed sensing theory, compressive spectral imagers aim to reconstruct the spectral images from compressive measurements using sparse signal recovery algorithms. In this paper, first, the image formation model and a sparsity-based reconstruction approach are presented for compressive photon-sieve spectral imager. Then the reconstruction performance of the approach is analyzed using different sparsity priors. In the system, a coded aperture is used for modulation and a photon-sieve for dispersion. In the measurements, coded and blurred images of spectral bands are superimposed. Simulation results show promising image reconstruction performance from these compressive measurements.

26th IEEE Signal Processing and Communications Applications Conference (SIU) -- MAY 02-05, 2018 -- Izmir, TURKEY WOS:000511448500298 This paper represents efficient expectation-maximization (EM) and space-alternating generalized expectation-maximization (SAGE) algorithm based iterative channel estimation methods for multi-cell massive multiple input multiple output (MIMO) systems. The proposed iterative channel estimation methods converge to the same mean square error (MSE) performance of the least squares (LS) estimator with the increasing number of iterations. LS channel estimation method requires conjugate transpose of a large-size pilot matrix. Conjugate transpose of the large-size matrix increases computational complexity. The size of the pilot matrix depends on the number of users. As the number of users increases, computational complexity increases too. In the proposed iterative channel estimation methods, the size of the pilot matrix is reduced to the vector size and the computational complexity is significantly reduced. The SAGE algorithm can estimate the channel by performing fewer iterations than EM, so it is a preferable method in terms of convergence speed. IEEE, Huawei, Aselsan, NETAS, IEEE Turkey Sect, IEEE Signal Proc Soc, IEEE Commun Soc, ViSRATEK, Adresgezgini, Rohde & Schwarz, Integrated Syst & Syst Design, Atilim Univ, Havelsan, Izmir Katip Celebi Univ

Sahne bölütleme ve etiketlemede son yıllarda süperpiksel tabanlı yaklaşımların başarılı olduğu görülmektedir. Öncelikle görüntü bir süperpiksel algoritması ile görsel olarak tutarlı küçük parçalara bölütlenmekte; daha sonra her süperpiksel için çeşitli öznitelik betimleyicileri hesaplanıp sınıflandırma yapılmaktadır. Bu bildiride süperpiksel öznitelikleri için Kernel Kodtablosu Kodlama (KKT) yaklaşımı önerilmiştir. KKT’de öznitelik vektörleri için sabit nicemleme yerine birden fazla kodsözcüğüne yumuşak eşlemleme yapılmaktadır. Kodsözcüklerine atanan ağırlıklar bir kernel uzaklık fonksiyonu ile hesaplanmaktadır. KKT yöntemi SuperParsing görüntü ayrıştırma algoritmasında SIFT özniteliklerinin kodlanması için kullanılmıştır. Geliştirdiğimiz yöntem 2,688 görüntü ve 33 sınıf içeren SIFT Flow veri kümesi kullanılarak denenmiş ve SuperParsing yönteminden %2.7 daha yüksek ayrıştırma doğruluğu sağlanmıştır. Superpixel based methods have recently shown success in scene segmentation and labeling. In scene labeling, a superpixel algorithm is used first to segment the image into visually consistent small regions; then several feature descriptors are computed and classification is performed for each superpixel. In this paper, Kernel Codebook Encoding (KCB) of superpixel features is proposed. In KCB feature vectors are mapped to multiple codewords in a soft manner, instead of the usual hard quantization. The weights assigned to the codewords are determined by a kernel distance function. KCB method is used for encoding of SIFT features in SuperParsing image parsing algorithm. The developed approach is tested on the SIFT Flow dataset consisting of 2,688 images and 33 classes, and achieves 2.7% increase in parsing accuracy over SuperParsing. Publisher's Version

Teknolojinin kullanımının yaygınlaşması ile beraber son zamanlarda insan-bilgisayar ilişkisi üzerine bir çok çalışma yapılmaktadır. Bilgisayar sistemlerinin insan eğitiminin üzerindeki etkisi ile ilgili yapılan çalışmalar buna örnek olarak gösterilebilir. Bu bildiride, hem oyun amaçlı hem de bilgisayar temelli eğitim programlarında kullanılan elektronik bateri oyunu için geliştirilen güvenilir ve gerçek zamanlı somut bir kullanıcı arayüz sistemi sunulmaktadır. Kullanıcı elektronik bagetleri kullanarak altı farklı yönde baget vuruşu gerçekleştirebilir. Geliştirilen darbe tespit algoritması, elektronik bagetlerden gelen sensör verilerini işleyerek hangi bagetten hangi yönde bir darbe hareketi oluştuğunu tespit eder. Bu bilgi ses sistemine gönderilerek ilgili bateri enstrümanına ait ses dosyasının çalınması sağlanır. Bu şekilde kullanıcıya gerçeğe yakın bir bateri deneyimi sunulmaktadır. Geliştirilen sistem, gerçek dünyada farklı kullanıcılar tarafından test edildi ve performansı değerlendirilmiştir. Elde edilen sonuçlar elektronik davul oyununun güvenilirliğini ve kullanılabilirliğini onaylamaktadır. Along with the widespread use of technology, there have been many studies on human-computer relation recently. Study of the effects of computer systems on human education can be shown as an example of this. In this paper, a reliable and real time tangible user interface system developed for the air drum game which is used both for game and computer based training programs is presented. A user can perform a play action in six different directions using electronic drum sticks. A developed algorithm detects the direction of moving stick by processing sensor data taken from these sticks. Information about the stick which is moved and its direction are sent to the sound system to play the sound file of the relevant drum instrument. In this way the user is presented with a realistic drum experience. The developed system has been tested by different users in real world and its performance has been reported. The results verify the reliability and usability of the electronic drum game. Publisher's Version

25th Signal Processing and Communications Applications Conference (SIU) -- MAY 15-18, 2017 -- Antalya, TURKEY, WOS: 000413813100148, Electrical harmonic is an important issue affecting power quality of electrical systems. Estimation of power harmonics are essential to design filters to be utilized in electrical power systems for reducing harmonics and their effects on the electrical power systems. In this study, a harmonic estimator based on differential evolution algorithm is proposed for harmonic estimation problems shown in power systems. Analyses of designed estimator are performed for two different experiments. In addition, performances of proposed estimator are compared with available results of the literature. The achieved results of this study showed that the proposed harmonic estimator in this paper is superior among previously reported ones., Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

5G teknolojilerinin olgunluk seviyelerinin yükseltilmesi donanım geliştirme faaliyetlerini gerektirir ve altyapının yeterliliği önem arz etmektedir. Bu bildiride, 5G araştırma sonuçlarının ürünleşmesine destek olmak amacıyla çok çekirdekli ARM-DSP yonga kümesi üzerinde geliştirilen 5G teknoloji geliştirme platformu sunulmaktadır. Önerilen platformda, fiziksel katman yazılımlarının gerçek zamanlı koşması için yüksek sinyal işleme yeteneğine sahip ve paralel çalışabilen çok çekirdekli DSP işlemciler kullanılırken üst katman yazılımları için üzerinde tümleşik işletim sistemlerinin çalıştırıldığı ARM işlemciler kullanılmaktadır. Önerilen mimarinin verimliliğini göstermek amacıyla 20 MHz bant genişliğinde çalışan fiziksel katman alıcı ve verici yazılımları çok çekirdekli DSP mimarisinde geliştirilmiş ve USRP radyolar ile gerçek zamanlı video transferi başarılı bir şekilde gösterilmiştir. 5G araştırma geliştirme faaliyetlerini desteklemek amacıyla bu platformun araştırmacıların hizmetine sunulması planlanmaktadır. Advancing the readiness level of 5G technologies requires hardware development activities, where the infrastructure has outmost importance. In this paper, we present multicore ARM-DSP based 5G technology development platform which will expedite commercializing the findings of the 5G researches. In the proposed platform, multicore DSP processors with enourmous signal processing capabilities are used to realize the real-time implementation of 5G physical layer algorithms while multicore ARM processors running real-time operating system are used to implement the real-time implementation of 5G higher layer protocols. To demonstrate the efficiency of the proposed platform, the proof of concept implementation of both transmitter and receiver algorithms within multicore DSP processors are performed for 20 MHz bandwidth and the real-time video transmission over the air using the USRP radios are successfully demonstrated. This platform will be made available to researchers to support 5G R&D activities.

UF-EMG test, in which non-invasive uroflowmetry (UF) and electromyography (EMG) signals are simultaneously recorded, is frequently used in children diagnosed with lower urinary tract dysfunction disease (AUSD) and its treatment. In the literature, independent (single) UF signals and integrated (dual) UF-EMG signals are graded many times but there is no classification study of UF-EMG integrated signals with Artificial Neural Networks (ANN), although studies have been done to classify UF signals with ANN. In this paper, it was aimed to classify the UF-EMG signals recorded from pediatric patients during the UF-EMG tests in Urodinami Center of Gulhane Education and Research Hospital using ANN. 773 (80%) of the 967 patients with an average age of 8 were used for training and 194 (20%) were used for the test. In YSA, the contribution of the features obtained from the EMG signals played a crucial role and was the main reason to improve the signal classification from 58% to 84.02%. The new classification method created by the obtained data does facilitate the interpretation of UF-EMG results for the clinical personnel in diagnosis, follow-up and treatment of patients. It is also aimed that the pediatric patients living in regions with less access to health care can be treated by providing an early and easy preliminary diagnostic tool., Günümüzde çocuklarda alt üriner sistem disfonksiyonu (AÜSD) hastalıklarının teşhisinde ve tedavinin takibinde, girişimsel olmayan (noninvaziv) üroflovmetri (UF) ve elektromiyografi (EMG) sinyallerinin eşzamanlı olarak kaydedildiği UF-EMG testi sıklıkla kullanılmaktadır. Literatürde, bağımsız UF sinyalleri ve UF-EMG sinyalleri birçok defa derecelendirilmiş, UF sinyallerinin Yapay Sinir Ağları (YSA) ile sınıflandırılmasına dair çalışmalar yapılmış olmasına rağmen, UF-EMG sinyallerinin birlikte kullanıldığı YSA ile sınıflandırma çalışması bulunmamaktadır. Bu çalışmada, Gülhane Eğitim ve Araştırma Hastanesi Ürodinami Merkezinde, pediatrik hastalara uygulanan UFEMG testleri sırasında kaydedilen sinyallerin YSA kullanılarak sınıflandırılması amaçlanmıştır. Yaş ortalaması 8 olan 967 hasta verisinden 773 adedi (%80) eğitim, 194 adedi (%20) ise test amaçlı kullanılmış, YSA’da EMG sinyallerinden elde edilen özniteliklerin katkısı başarı oranının %58‘den %84,02 seviyelerine gelmesini sağlamıştır. Elde edilen veriler ile oluşturulan yeni sınıflandırma metodu hastaların tanı, takip ve tedavisinde sağlık personelinin UF-EMG sonuçlarını yorumlamasında kolaylık sağlamayı; sağlık hizmetine ulaşma imkânı daha az olan bölgelerdeki çocuk hastalara, daha erken ve kolay ön tanı konularak uygun tedaviye yönlendirme imkânı sağlamayı hedeflemektedir.

Karakaş, Sirel (Dogus Author) -- Doğutepe, Elvin (Dogus Author) -- Conference full title : 25th Signal Processing and Communications Applications Conference (SIU); 15 May 2017 through 18 May 2017. Antalya; Turkey Dikkat Eksikliği ve Hiperaktivite Bozukluğu (DEHB), psikiyatri servislerine başvuran çocuklarda en çok rastlanan rahatsızlıktır. Bu bozukluk 20. Yüzyılın başında tanımlanmasına rağmen, teşhisinde halen ciddi sorunlar yaşanmaktadır. Bu çalışmada, Olaya ilişkin Potansiyel (OİP) verilerinden çıkarılan zaman frekans özellikleri kullanılarak DEHB olan çocukların bu bozukluğu taşımayanlara göre sınıflandırılması elde edilmiştir. Bu özniteliklerin DEHB teşhisinde en etkili olanları Destek Vektör Makinesi – Yinelemeli Öznitelik Seçimi yöntemiyle belirlenmiş ve bu özniteliklerle eğitilen Destek Vektör Makinesi, üç öznitelik kullanıldığında %98 ve altı öznitelik kullanıldığında %99.5 doğrulukla DEHB teşhisi koymayı başarmıştır. Sonuçlar Tekrarlanan Çapraz Geçerlilik Yöntemiyle elde edilmiştir. Ayrıca, bu yöntem Kararlılık Seçimi ile desteklenerek DEHB tanısında en etkili öznitelikler belirlenmiştir. Ayrıca yeni bir tasarımla önerilen düzenek test edilmiş ve % 100 doğruluk elde edilmiştir. Sonuçlar, çalışmadaki düzeneğin psikiyatrik bozuklukların tanısında ve seyrinde kullanılabileceğini göstermektedir. Attention-deficit/hyperactivity disorder (ADHD) is the most frequent diagnosis among children who are referred to psychiatry departments. Although ADHD was discovered at the beginning of the 20th century, its diagnosis is confronted with many problems. In this paper, a novel classification approach that discriminates ADHD and non-ADHD groups over the timefrequency domain features of ERP recordings is presented. Support Vector Machine-Recursive Feature Elimination (SVMRFE) was used to obtain best discriminating features. When only three of these features were used the accuracy of classification reached to 98%, and use of six features further improved classification accuracy to 99.5%. The proposed scheme was tested with a new experimental setup and 100% accuracy is obtained. The results were obtained using RCV. The classification performance of this study suggests that TFHA can be employed as a core component of the diagnostic and prognostic procedures of various psychiatric illnesses.

In this paper, we first form a sensor using low power and low cost components, that is, MSP430G2553 microcontroller, nRF24L01+ communication unit, and a Light Dependent Resistor (LDR). Then, by communicating such sensors with a Fusion Center (FC), we form a Wireless Sensor Network (WSN). Sensors measure the light intensities at their locations and send their measurements to the FC. Then, FC forwards the sensor measurements to the computer which performs field estimation using ordinary kriging. Our test results show that the predicted field measurement using ordinary kriging at a specific location is quite close to the actual sensor measurement at that location. Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

Due to copyright restrictions, the access to the full text of this article is only available via subscription. Bu çalışmada, görünür ışıkla haberleşme (VLC) ağları için genetik algoritma (GA) temelli kaynak atama tekniği önerilmiştir. Optik VLC agı mevcut aydınlatma altyapısının kullanıldığı iç mekanlarda çok küçük boyutlu hücrelerden oluşur. Sistem kapasitesini artırmak ve kaynakları kullanıcılar arasında daha adaletli bir şekilde dağıtmak için VLC ağlarında etkili kaynak atama mekanizmalarına gerek vardır. Bu nedenle, bu çalışmada GA temelli kaynak atama tekniği önerilerek toplam kanal kapasitesini maksimize etme amaçlanmıştır. Önerilen kaynak atama algoritması Round Robin ve Best CQI algoritmalarına göre sistem kapasitesini arttırma açısından daha iyi sonuçlar vermiştir. In this paper, we present a genetic algorithm (GA) based resource allocation technique for indoor optical visible light communication (VLC) networks. Optical VLC network using existing lighting infrastructure in an indoor environment is a cellular network with very small size cells. It is necessary to have effective resource allocation mechanisms for VLC networks to maximize system throughput and fairness among users. We propose a GA based resource allocation scheme to improve total system throughput. Our proposed scheme gives better results compared to Round Robin and Best CQI algorithms in terms of system throughput. TÜBİTAK

25th Signal Processing and Communications Applications Conference (SIU) -- MAY 15-18, 2017 -- Antalya, TURKEY WOS: 000413813100575 Large scale dictionaries such as wordnets are one of the important structures used in natural language processing. Wordnet is a comprehensive dictionary containing semantic relations including synonyms, antonyms, etc. In this paper, we try to extract Hypernym-Hyponym relations which are one of the main parts of WordNet. For this aim, we used Wikipedia, Contemporary Dictionary of Turkish and Wiktionary as corpora. We used Finite State Machines for developing several rules for Hypernym-Hyponym extraction. Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

25th Signal Processing and Communications Applications Conference (SIU) -- MAY 15-18, 2017 -- Antalya, TURKEY WOS: 000413813100213 Superpixel based methods have recently shown success in scene segmentation and labeling. In scene labeling, a superpixel algorithm is used first to segment the image into visually consistent small regions; then several feature descriptors are computed and classification is performed for each superpixel. In this paper, Kernel Codebook Encoding (KCB) of superpixel features is proposed. In KCB feature vectors are mapped to multiple codewords in a soft manner, instead of the usual hard quantization. The weights assigned to the codewords are determined by a kernel distance function. KCB method is used for encoding of SIFT features in SuperParsing image parsing algorithm. The developed approach is tested on the SIFT Flow dataset consisting of 2,688 images and 33 classes, and achieves 2.7% increase in parsing accuracy over SuperParsing. Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

25th Signal Processing and Communications Applications Conference (SIU) -- MAY 15-18, 2017 -- Antalya, TURKEY, WOS: 000413813100403, In this paper, a spectral graph based image coding method is proposed. In the proposed method, only the smoothing parameter in the graph adjacency matrix, graph Fourier transform (GFT) coefficients kept after hard thresholding and column indices of eigenvectors corresponding to kept coefficients are sent to the decoder side. It is advantageous in that this coding process is performed by sending much less information as compared to other graph based methods. Our results indicate that our proposed method enables image compression without deteriorating the visual quality by using very small numbers of GFT coefficients., Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

25th Signal Processing and Communications Applications Conference (SIU) -- MAY 15-18, 2017 -- Antalya, TURKEY WOS: 000413813100051 In this paper, a wavelet based power transformer protection algorithm, which is one of the most important components of power system, is proposed. Traditionally, power transformers (rated power greater than 1.5MVA) are protected by differential protection scheme. Its function supported by discrete Fourier transform is fundamentally based on comparison of primary and secondary currents of the transformer and dedicated to discriminate internal faults and magnetizing inrush phenomena. In this way, the relay (differential protection relay) is ensured to be more reliable and selective. Nevertheless, the concept of transformer protection has been keeping up-to-date and several numerical protection algorithms have been suggested for this purpose. In this work, maximal overlap discrete wavelet transform (MODWT) based transformer differential protection technique is proposed. The features vectors obtained by MODWT are then processed by artificial neural network (ANN) to classify the normal, magnetizing, and internal fault conditions. The obtained results are also compared to traditional discrete wavelet transform (DWT) and interpreted using performance curves. Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

21st National Biomedical Engineering Meeting (BIYOMUT) -- NOV 24-DEC 26, 2017 -- Istanbul, TURKEY WOS: 000447671500032 In pharmacological experiments, the effect of ion concentrations on Action potential (AP) and its association with important neurological disorders such as stroke, epilepsy and depression has been investigated. In this paper, Hodgkin-Huxley (HH) and Morris-Lecar (ML) membrane models, which have contributed to the literature of theoretical neural science, have been studied. In the first step, these two models have been simulated individually in Simulink environment using Matlab software. In the second step, the ratio of the amount of extracellular ion to the amount of intracellular ion was expressed as Hd/Hi. Based on Na. and K. as ions, the effect of Hd/Hi ratio change on AP signal was investigated based on HH and ML as a model. In the HH model, it was searched that the Hd/Hi change due to sodium ions influenced the amplitude of AP and the depolarization phase. The change in Hd/Hi ratio due to potassium ion concentrations was found to affect the threshold and hyper-polarization phase of AP. In the ML membrane model, it was observed that the change in Hd/Hi ratio due to intracellular and extracellular Ca+2 and K- ion concentrations affected the formation time of AP, firing time, amplitude and formation phases. HH and ML models showed similarity in the same parameters.

25th Signal Processing and Communications Applications Conference, SIU 2017; Antalya; Turkey; 15 May 2017 through 18 May 2017, Kompozit malzemelerde imalat kalitesini arttırabilmek için, malzeme davranışının tüm üretim aşamalarında izlenmesi gerekmektedir. Bu amaçla fiber optik sensörlerin kullanılması, özellikle de fiber Bragg ızgaraların (Fiber Bragg Grating, FBG) kompozit malzemelere yerleştirilmesi hızla yaygınlaşmaktadır. Bu bağlamda çalışmamızda, FBG tabanlı alternatif bir sensör sorgulama düzeneği tasarlanmış ve kompozit malzemelerin üretimi sırasında ortaya çıkan kalıcı (residual) streslerin tespit edilmesi yönündeki teorik analizleri tamamlanmıştır. Makale kapsamında ayrıca fiber takviyeli termoset kompozit malzemelere FBG sensörler yerleştirilerek üretim süreçlerinin gözetimi yönünde deneysel ön-hazırlıklar gerçekleştirilmiştir., It is essential to monitor the behavior of composite materials during manufacturing process to ensure the high quality of manufactured materials. For this purpose, the use of optical fiber sensing, particularly the embedding of fiber Bragg grating (FBG) sensors into composite materials has been gaining growing popularity. In this context, we design a FBG-based alternative sensor interrogation system and completed the theoretical analysis for the prospect of providing the residual stresses appearing during fabrication process of composite materials. It has been also realized, in this paper, the preliminary experimental trials by monitoring FBG sensors embedded into fiber reinforced thermo-set composites for the purpose of process monitoring.

25th Signal Processing and Communications Applications Conference, SIU 2017; Antalya; Turkey; 15 May 2017 through 18 May 2017, Manzara metni tespit etme ve tanıma gerçek dünya imgelerinden anlamlı içerik çıkarmada ve bu imgelerden olu-şan büyük veri kümelerinde aranan metni bulup getirmede büyük kolaylık sağladığından araştırmacılardan artan bir ilgi görmektedir. Buna rağmen, bu alanda yaygın olarak kullanılan imge kümelerinin büyük kısmı İngilizce metinler içerdiklerinden yapılan çalışmalar tek dil ile kısıtlı kalmıştır. Buradan yola çıkarak, geliştirilen manzara metni tespit etme ve tanıma teknolojilerinin Türkçe metinlere uygulanması, performanslarının analiz edilmesi ve Türkçe’ye özgün algoritmaların geliştirilmesi amacıyla literatürde ilk kez bir Türkçe manzara metni veri kümesi oluşturulmuştur. Bu bildiride, kısaca STRIT (Scene Text Recognition In Turkish) olarak adlandırılan veri kümesinin içeriği anlatılmaktadır. Buna ek olarak, Türkçe manzara tespiti ve tanınması için iki taban çizgisi yöntemi denenmekte ve ön sonuçlar takdim edilmektedir., Scene text localization and recognition keeps attracting an increasing interest from researchers due to its valuable advantage in extracting content from real world images and in image retrieval via text search. Nevertheless, due to the fact that the majority of the image datasets that are commonly used in this field is comprised of text in English, the related studies have mostly been limited to a single language. On that account, in order to apply the technologies developed for scene text detection and recognition to Turkish scene text, analyze their performances and to develop Turkish language specific algorithms, a Turkish scene text database is collected for the first time in the literature. In this paper, the contents of this database, shortly called STRIT (Scene Text Recognition In Turkish), are detailed. Additionally, two baseline methods are tested to detect and recognize scene text in Turkish and the preliminary results are presented., TUBITAK (BIDEB-114C025)

25th Signal Processing and Communications Applications Conference (SIU) -- MAY 15-18, 2017 -- Antalya, TURKEY WOS: 000413813100519 Along with the widespread use of technology, there have been many studies on human-computer relation recently. Study of the effects of computer systems on human education can be shown as an example of this. In this paper, a reliable and real time tangible user interface system developed for the air drum game which is used both for game and computer based training programs is presented. A user can perform a play action in six different directions using electronic drum sticks. A developed algorithm detects the direction of moving stick by processing sensor data taken from these sticks. Information about the stick which is moved and its direction are sent to the sound system to play the sound file of the relevant drum instrument. In this way the user is presented with a realistic drum experience. The developed system has been tested by different users in real world and its performance has been reported. The results verify the reliability and usability of the electronic drum game. Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

Medical Technologies National Congress (TIPTEKNO) -- OCT 12-14, 2017 -- TRABZON, TURKEY, WOS: 000427649500046, In this paper, the effects of autapse on the first spike latency of the stochastic H-H neuron are examined. In the study, it is considered that H-H neuron has an electrical autapse and by applying a suprathreshold periodic signal to neuron the first spike times has been observed. Obtained results show that the first spike latency of H-H neuron increases prominently in a certain autaptic time delay with the increasing of autaptic conductance. Also, the first spike latency decreases with the increasing of autaptic conductance in a different autaptic time delay interval. In the context of these results, we come to conclusion that the autapse have played important roles on the control of first spike latency of stochastic H-H neurons., IEEE Turkey Sect

25th Signal Processing and Communications Applications Conference (SIU) -- MAY 15-18, 2017 -- Antalya, TURKEY, WOS: 000413813100592, In this paper, a study which automatically detects four open string notes of the violin from the recorded violin sounds is presented for the purpose of using in the beginner level violin education. First, in order to be able to analyze the violin sound recordings, it is necessary for the music piece to be segmented into the notes forming the music piece itself For this, it is necessary to correctly determine the starting times (onsets) of the notes. Because of the envelope characteristic that the violin sound has, it is more difficult to detect the note onset than other musical instrument sounds such as piano and guitar. By using the complex spectral difference method, the note onsets were determined, then by determining the fundamental frequency the note of the violin sound was found. The proposed system was applied to the open string sound tracks obtained from eight people by using the iPad-based professional recording system, and the results showed that the proposed system worked successfully., Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

25th Signal Processing and Communications Applications Conference (SIU) -- MAY 15-18, 2017 -- Antalya, TURKEY, WOS: 000413813100200, In this paper, the performance of direct current informative subcarrier intensity modulation (DCI-SIM) which is proposed to enhance the power efficiency of subcarrier intensity modulation (SIM) is investigated in visible light communication. The BER performance of DCI-SIM is compared with SIM, pulse amplitude modulation (PAM) and, enhanced unipolar orthogonal frequency division multiplexing (eU-OFDM) for different spectral efficiencies. Using analytical derivations and simulation results, we show that the DCI-SIM is the most power efficient modulation technique among others, and that the power efficiency increases with the spectral efficiency. Furthermore, we optimize the constellation of 64-DCI-SIM and present the results., Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

25th Signal Processing and Communications Applications Conference, SIU 2017; Antalya; Turkey; 15 May 2017 through 18 May 2017, Bu çalışmada harflerin yazılma hareketlerinden yararlanılarak el yazısı karakter tanıma işlemi yapılmıştır. Yazma hareket bilgisini elde etmek adına 3 yönlü ivmeölçer kullanılmıştır. Mevcut çoğu sensörde olduğu gibi, ivmeölçerlerde istenen ivme bilgisinin yanında istenmeyen gürültü bilgisi de elde edilmektedir. Bu nedenle tanıma algoritmasından önce sinyaller ön işlemeden geçirilmeli ve gürültüden arındırılıp tanıma için uygun hale getirilmelidir. Bu amaca uygun olarak sinyallere Deneysel Mod Ayrıştırması (EMD) ve normalizasyon ön işleme adımları uygulanmıştır. Daha sonra sinyaller Dinamik Zaman Bükme (DTW) metodu ile karşılaştırılmış ve %91.92’lik doğru sınıflandırma oranı elde edilmiştir., In this paper, handwritten character recognition by using characters' writing movements is investigated. To obtain the information about writing movements a 3-axis accelerometer is used. Just like most of other sensors, 3-axis accelerometers give the actual movement signal as well as noise. Before the recognition step, all of the signals need to be preprocessed and the noisy parts need to be removed. So, Empirical Mode Decomposition (EMD) and normalization preprocessing steps are applied to the signals. Finally, the signals in the dataset are compared with Dynamic Time Warping for classification and accurate classification rate of 91.92% is obtained.

Due to copyright restrictions, the access to the full text of this article is only available via subscription. Bu bildiride çok-girdili çok-çıktılı (multiple-input multiple-output, MIMO) ve ortogonal frekans bölmeli çoklu erişim (orthogonal frequency division multiple access, OFDMA) tabanlı hücresel kablosuz ağlarda farklı algoritmalara dayalı olarak birçok kullanıcıya, kaynakları tahsis eden zamanlayıcıların performansları incelenmektedir. OFDMA’in alışılmış zaman ve frekans kaynaklarına ek olarak, MIMO yapısı ile oluşan anten kaynakları, algoritmalara dahil edilmiştir. MIMO ve OFDM teknolojileri hakkında temel bilgiler verildikten sonra, Long Term Evolution (LTE) aglarında Round Robin (RR), Proportional Fair (PF) ve Maximum Throughput (MT) zamanlayıcılarının performsları değerlendirilmiştir. Makro hücre kentsel alan kanal modeli kullanılarak, çoklu anten yapısı içeren LTE sistem mimarisi dikkate alınmıştır. Performans değerlendirmeleri için Shannon Kapasitesi (Shannon Capacity) ve Jain’in adillik endeksi (Jain’s fairness index) kullanılmıştır. Buna ek olarak, kapasiteyi en üst düzeye çıkarma ile adilliğe ulaşma arasında bulunan ödünleşmeler Monte Carlo simülasyonlarıyla gösterilmiştir. In this paper, we investigate the performances of schedulers, which basically allocate the resources to multiple users based on different algorithms, in multiple-input multiple-output (MIMO) and orthogonal frequency division multiple access (OFDMA) based cellular wireless networks. In addition to conventional OFDMA resources which are time and frequency, antennas are also considered as resource during scheduling process. After explaining the basic concepts of MIMO and OFDMA technologies, the performances of Round Robin (RR), Proportional Fair (PF) and Maximum Throughput (MT) scheduling algorithms are evaluated in Long Term Evolution (LTE) networks. LTE system architecture including multiple antenna structure with the use of macro cell urban area channel model is considered. Shannon capacity and Jain's fairness index are utilized for performance evaluations. Additionally, the common trade-offs between maximizing capacity and achieving fairness are presented through Monte Carlo simulations.

In this paper, a new power control algorithm based on busy tone approach has been proposed for the coordination of IEEE 802.22 and IEEE 802.11af systems in TV white space. Different from the earlier studies, in addition to both 802.11af access point and clients listening to the busy tone, they also adjust their communication power according to the location information and use hopping for communication, if needed Acccordingly, interference caused to 802.22 systems has been reduced while the 802.11af systems are still able to communicate. This study quantifies the 802.11af and 802.22 system performances in terms of interfering packet rate and succesful packet transmission rate for different scenarios considering the communication parameters and channel models adapted for the standards. Turk Telekom Arcelik A S Aselsan ARGENIT HAVELSAN NETAS; Adresgezgini IEEE Turkey Sect AVCR Informat Technologies Cisco i2i Syst Integrated Syst & Syst Design ENOVAS FiGES Engn MS Spektral Istanbul Teknik Univ

Due to copyright restrictions, the access to the full text of this article is only available via subscription. Yazılım-Tanımlı Veri Depolama (Software-Defined Storage, SDS) konsepti son zamanlarda veri merkezlerinde yeni bir teknoloji olarak ortaya çıkmıştır. Bu teknoloji, hem veri merkezi sahiplerine, hem uygulama geliştiricilerine hem de Mobil Ağ Operatörleri (Mobile Network Operator, MNO), Mobil Sanal Ağ Operatörleri (Mobile Virtual Network Operator, MVNO) vb. dahil olmak üzere abonelere yüksek kullanılabilirlik, yönetilebilirlik, performans ve maliyet avantajları sağlayabilir. Bu çalışmada MNO’ların veri merkezindeki depolama birimlerine erişimini düzenleyen Max-Min Fairness programcı algoritmasının performansı, veri merkezindeki talepler ve mevcut depolama kaynakları temel alınarak incelenmektedir. Kaynakların tahsisi, garanti edilen ve anlık depolama miktarı göz önünde bulundurularak yapılmaktadır. Performans göstergesi (key performance indicator, KPI) olarak doymuş-MNO-oranı (satisfied-MNO-ratio) ve en basit tahsis algoritması olan, Round Robin, karşılaştırmalarda kıstas olarak kullanılmaktadır. Software-Defined Storage (SDS) concept has recently emerged as a new technology inside data centers. It can provide high availability, manageability, performance and cost benefits to both data center owners, application developers and subscribers including Mobile Network Operators (MNO) or Mobile Virtual Network Operators (MVNO). In this paper, based on the assumption that MNOs are competing to access the storage units inside data centers, performance of Max-Min Fairness scheduler algorithms is investigated based on MNO's demands and available storage resources inside data center. The allocation of resources are performed under the consideration of both guaranteed and instantaneous storage amount. Satisfied-MNO-ratios are used as key performance indicator (KPI) and simplest allocation algorithm, Round Robin, is considered as benchmark in comparison. http://ieeexplore.ieee.org/document/7960313/

25th Signal Processing and Communications Applications Conference (SIU) -- MAY 15-18, 2017 -- Antalya, TURKEY WOS: 000413813100251 The Internet and social media provide a major source of information about people's opinions. Due to the rapidly growing number of online documents, it becomes both time-consuming and hard task to obtain and analyze the desired opinionated information. Sentiment analysis is the classification of sentiments expressed in documents. To improve classification perfromance feature selection methods which help to identify the most valuable features are generally applied. In this paper, we compare the performance of four feature selection methods namely Chi-square, Information Gain, Query Expansion Ranking, and Ant Colony Optimization using Maximum Entropi Modeling classification algorithm over Turkish Twitter dataset. Therefore, the effects of feature selection methods over the performance of sentiment analysis of Turkish Twitter data are evaluated. Experimental results show that Query Expansion Ranking and Ant Colony Optimization methods outperform other traditional feature selection methods for sentiment analysis. Turk Telekom, Arcelik A S, Aselsan, ARGENIT, HAVELSAN, NETAS, Adresgezgini, IEEE Turkey Sect, AVCR Informat Technologies, Cisco, i2i Syst, Integrated Syst & Syst Design, ENOVAS, FiGES Engn, MS Spektral, Istanbul Teknik Univ

Date of Conference: 15-18 May 2017 Conference Name: IEEE 25th Signal Processing and Communications Applications Conference, SIU 2017 In this paper, a novel method for extracting radar fingerprint using the unintentional modulation on radar signals is proposed. Proposed technique decomposes the unintentional modulations into its components using Variational Mode Decomposition (VMD) technique. Then, features that characterize each component are calculated. Simulations using real radar data show that proposed technique can classify radars in the dataset with high performance.