Browsing by Author "Öz, Dindar"

Now showing 1 - 4 of 4

Göç Eden Kuşlar Algoritmasinda Kaos Fonksiyonlarinin Kullanilmasi
(2016) DİNDAR ÖZ; Öz, Dindar
Olasılıksal eniyileme algoritmaları çalışmalarının birçok aşamasında rastlantısal veri kullanmaktadırlar ve performansları büyük oranda bu rastlantısal verinin dağılımına göre değişiklik göstermektedir. Bu noktadan hareketle farklı rastlantısal veri kaynaklarının eniyileme algoritmalarının performansına etkisi son zamanlardaki birçok çalışmanın odak noktası olmuştur. Kaotik eşlem fonksiyonları matematiksel özellikleri sonucu rastlantısal veri kaynağı olarak kullanılmaya oldukça elverişlidir. Bu çalışmada kaotik eşlem fonksiyonlarının popülasyon tabanlı evrimsel bir algoritma olan göç eden kuşlar algoritmasına etkisi bilgisayar mimarisinin güncel problemlerinden biri olan görev dağıtım problemi üzerinde deneysel olarak incelenmiştir. Deneyler neticesinde bir kısım kaotik eşlem fonksiyonlarının ele alınan problem için uygun olmadığı gözlense de klasik rastlantısal veri üretme algoritmaları ile başa baş performans sergileyen kaotik eşlem fonksiyonlarının da bulunduğu görülmüştür
Citation - WoS: 31
Citation - Scopus: 43
Optimal Any-Angle Path finding In Practice
(AI ACCESS FOUNDATION, 2016) Daniel Harabor; Alban Grastien; Dindar Oz; Vural Aksakalli; Harabor, Daniel; Öz, Dindar; Grastien, Alban; Aksakalli, Vural
Any-angle path finding is a fundamental problem in robotics and computer games. The goal is to find a shortest path between a pair of points on a grid map such that the path is not artificially constrained to the points of the grid. Prior research has focused on approximate online solutions. A number of exact methods exist but they all require super-linear space and pre-processing time. In this study we describe Anya: a new and optimal any-angle path finding algorithm. Where other works find approximate any-angle paths by searching over individual points from the grid Anya finds optimal paths by searching over sets of states represented as intervals. Each interval is identified on-the-fly. From each interval Anya selects a single representative point that it uses to compute an admissible cost estimate for the entire set. Anya always returns an optimal path if one exists. Moreover it does so without any offline pre-processing or the introduction of additional memory overheads. In a range of empirical comparisons we show that Anya is competitive with several recent (sub-optimal) online and pre-processing based techniques and is up to an order of magnitude faster than the most common benchmark algorithm a grid-based implementation of A*.
Sanal makine yerleştirme problemi için çok amaçlı optimizasyon çözümü
(2024) Altuntaş, Tolga Buğra; Öz, Dindar
Delivering different services over the Internet requires cloud computing. These services are managed by Cloud Service Providers using Virtual Machines that simulate physical machines in order to provide the required computing resources. Efficiently managing and allocating these resources is crucial for achieving optimal performance and cost-effectiveness. Nevertheless, the rapid expansion of cloud computing increased the complexity and scale of cloud environments. A key element of cloud computing is virtual machine placement (VMP), which makes sure that virtual machines are distributed among physical servers as efficiently as possible. Effective VMP strategies optimize data center performance and energy management, affecting operational cost and customer satisfaction. This work focuses on resource utilization and energy management on the multi-objective VMP problem. Extended Adapted Large Neighborhood Search (EALNS) algorithm is utilized to solve the problem. The EALNS algorithm uses a weight value to improve the spread of non-dominated solutions and create a better Pareto front. Five problem-specific destroy and repair operators are employed to adapt the EALNS algorithm to the VMP problem. To the best of our knowledge, this is the first work that uses the ALNS algorithm to solve a multi-objective VMP problem. The comparison experiments are done against three state-of-the-art multi-objective algorithms. The results show that the EALNS algorithm has great scalability and creates higher-quality Pareto fronts than its competitors.
Using Proxy Design Pattern for Transparent Redundant Execution
(Ceur-Ws, 2018) Öz, Sinan; Öz, Işıl; Öz, Dindar