8 sonuçlar
Arama Sonuçları
Listeleniyor 1 - 8 / 8
Yayın Spectral renormalization group for the Gaussian model and 𝜓4 theory on nonspatial networks(American Physical Society, 2015-08-06) Tuncer, Aslı; Erzan, AyşeWe implement the spectral renormalization group on different deterministic nonspatial networks without translational invariance. We calculate the thermodynamic critical exponents for the Gaussian model on the Cayley tree and the diamond lattice and find that they are functions of the spectral dimension, (d) over tilde. The results are shown to be consistent with those from exact summation and finite-size scaling approaches. At (d) over tilde = 2, the lower critical dimension for the Ising universality class, the Gaussian fixed point is stable with respect to a psi(4) perturbation up to second order. However, on generalized diamond lattices, non-Gaussian fixed points arise for 2 < <(d)over tilde> < 4.Yayın Network synchronization: Spectral versus statistical properties(Elsevier B.V., 2006-12) Atay, Fatihcan Mehmet; Bıyıkoğlu, Türker; Jost, JürgenWe consider synchronization of weighted networks, possibly with asymmetrical connections. Focusing on causal relations rather than the observed correlations, we show that the synchronizability of networks cannot be directly inferred from their statistical properties. Small local changes in the network structure can sensitively affect the eigenvalues relevant for synchronization, while the gross statistical network properties remain essentially unchanged. Consequently, commonly used statistical properties, including the degree distribution, degree homogeneity, average degree, average distance, degree correlation and clustering coefficient, can fail to characterize the synchronizability of networks in terms of causal relations, despite the observed correlations.Yayın Discovering cis-regulatory modules by optimizing barbecues(Elsevier Science Bv, 2009-05-28) Mosig, Axel; Bıyıkoğlu, Türker; Prohaska, Sonja J.; Stadler, Peter F.Gene expression in eukaryotic cells is regulated by a complex network of interactions, in which transcription factors and their binding sites on the genomic DNA play a determining role. As transcription factors rarely, if ever, act in isolation, binding sites of interacting factors are typically arranged in close proximity forming so-called cis-regulatory modules. Even when the individual binding sites are known, module discovery remains a hard combinatorial problem, which we formalize here as the Best Barbecue Problem. It asks for simultaneously stabbing a maximum number of differently colored intervals from K arrangements of colored intervals. This geometric problem turns out to be an elementary, yet previously unstudied combinatorial optimization problem of detecting common edges in a family of hypergraphs, a decision version of which we show here to be NP-complete. Due to its relevance in biological applications, we propose algorithmic variations that are suitable for the analysis of real data sets comprising either many sequences or many binding sites. Being based on set systems induced by interval arrangements, our problem setting generalizes to discovering patterns of co-localized itemsets in non-sequential objects that consist of corresponding arrangements or induce set systems of co-localized items. In fact, our optimization problem is a generalization of the popular concept of frequent itemset mining.Yayın Improving the calibration time of traffic simulation models using parallel computing technique(Institute of Electrical and Electronics Engineers Inc., 2019-06) Dadashzadeh, Nima; Ergün, Murat; Kesten, Ali Sercan; Zura, MarijanThe calibration procedure for traffic simulation models can be a very time-consuming process in the case of a large-scale and complex network. In the application of Evolutionary Algorithms (EA) such as Genetic Algorithms (GA) and Particle Swarm Optimization (PSO) for calibration of traffic simulation models, objective function evaluation is the most time-consuming step in such calibration problems, because EA has to run a traffic simulation and calculate its corresponding objective function value once for each set of parameters. The main contribution of this study has been to develop a quick calibration procedure for the parameters of driving behavior models using EA and parallel computing techniques (PCTs). The proposed method was coded and implemented in a microscopic traffic simulation software. Two scenarios with/without PCT were analyzed using the developed methodology. The results of scenario analysis show that using an integrated calibration and PCT can reduce the total computational time of the optimization process significantly-in our experiments by 50%-and improve the optimization algorithm's performance in a complex optimization problem. The proposed method is useful for overcoming the limitation of computational time of the existing calibration methods and can be applied to various EAs and traffic simulation software.Yayın Multiband matching network design via transformation based real frequency approach(IEEE, 2016) Yıldız, Serkan; Aksen, Ahmet; Yarman, Bekir Sıddık BinboğaIn this paper, a new method is presented for multiband matching network design. Low pass to band pass (LP-BP) frequency transformation is incorporated with the parametric Real Frequency Technique (RFT) for creating multiband network functions. In the new parametric approach, the poles of a low pass type impedance function are optimized under LP-BP transformation, for multiband matching of complex terminations. The resultant matching network is realized with resonance sections yielding multiband transducer power gain (TPG) characteristic.Yayın Statistical analysis of bus transportation networks of Istanbul(World Scientific Publishing Co Pte Ltd, 2016) Çoban, Veysel; Atan, Sabri TankutTransportation networks such as railway, airport and bus networks are the real-world networks whose inherent statistical properties characterize and differentiate the networks. In order to understand the network characteristics of bus transportation networks (BTNs) of Istanbul, we analyzed its network properties such as degree distributions, clustering coefficients and assortativity. BTNs of Istanbul is defined into three networks as the existence and nonexistence of the metrobus and existence of third- bridge. They are also graphically represented within C-, L- and P-Space topologies that are defined with the connection of the bus stops or routes. Statistical results obtained from network properties reflected the characteristics of the BTNs of Istanbul and give an information about the effects of the metrobus lines and third bridge on the BTNs in Istanbul.Yayın Richards uzayında band geçiren devre fonksiyonu gerçeklemesi ve yama anten uyumlaştırmada kullanımı(IEEE, 2014-04-23) Köprü, Ramazan; Aydın, Çağatay; Yarman, Bekir Sıddık BinboğaLiteratürde çok iyi bilinmektedir ki, Richards-düzlemi, Laplace-düzleminde tanjant hiperbolik eşlemesi uygulanarak elde edilen dönüştürülmüş bir uzaydır. Richards frekansı cinsinden üretilen devre fonksiyonları gerçek frekans ekseninde π periyoduna sahiptir. Richards uzayında alçak geçiren prototip devre fonksiyonu tasarlandığında, bu periyodik özellik nedeniyle, karşı gelen periyodik band geçiren devre fonksiyonu frekans ekseninde belirli bandlarda tekrarlanarak ortaya çıkmaktadır. Tasarımcı, uygulamanın gereksinimlerine göre bu tekrarlanan bandlar arasından ilgilendiği bandı seçebilir. Bu çalışmada, 3.2448-3.744 GHz bandında çalıştırılmak üzere tasarlanmış UWB uygulamaları için elverişli bir mikroşerit yama anten için eş-uzunluklu (commensurate) iletim hatları ile oluşturulmuş uyumlaştırma devresi tasarımı ele alınmaktadır. Tasarımda, SRFT (Simplified Real Frequency Technique: Basitleştirilmiş Gerçel Frekans Tekniği) kullanılmaktadır ve teorik tasarım ile MWO (AWR) benzeşimleri arasında çok iyi uyum olduğu gözlenmiştir.Yayın Adaptive locally connected recurrent unit (ALCRU)(Springer Science and Business Media Deutschland GmbH, 2025-07-03) Özçelik, Şuayb Talha; Tek, Faik BorayResearch has shown that adaptive locally connected neurons outperform their fully connected (dense) counterparts, motivating this study on the development of the Adaptive Locally Connected Recurrent Unit (ALCRU). ALCRU modifies the Simple Recurrent Neuron Model (SimpleRNN) by incorporating spatial coordinate spaces for input and hidden state vectors, facilitating the learning of parametric local receptive fields. These modifications add four trainable parameters per neuron, resulting in a minor increase in computational complexity. ALCRU is implemented using standard frameworks and trained with back-propagation-based optimizers. We evaluate the performance of ALCRU using diverse benchmark datasets, including IMDb for sentiment analysis, AdditionRNN for sequence modelling, and the Weather dataset for time-series forecasting. Results show that ALCRU achieves accuracy and loss metrics comparable to GRU and LSTM while consistently outperforming SimpleRNN. In particular, experiments with longer sequence lengths on AdditionRNN and increased input dimensions on IMDb highlight ALCRU’s superior scalability and efficiency in processing complex data sequences. In terms of computational efficiency, ALCRU demonstrates a considerable speed advantage over gated models like LSTM and GRU, though it is slower than SimpleRNN. These findings suggest that adaptive local connectivity enhances both the accuracy and efficiency of recurrent neural networks, offering a promising alternative to standard architectures.
