Browsing by Subject "Genetic algorithm"

Now showing 1 - 15 of 15

Open Access
A utilization based genetic algorithm for virtual machine placement in cloud systems
(2024-01-15) Çavdar, Mustafa Can; Körpeoğlu, İbrahim; Ulusoy, Özgür
Due to the increasing demand for cloud computing and related services, cloud providers need to come up with methods and mechanisms that increase the performance, availability and reliability of data centers and cloud systems. Server virtualization is a key component to achieve this, which enables sharing of resources of a single physical machine among multiple virtual machines in a totally isolated manner. Optimizing virtualization has a very significant effect on the overall performance of a cloud computing system. This requires efficient and effective placement of virtual machines into physical machines. Since this is an optimization problem that involves multiple constraints and objectives, we propose a method based on genetic algorithms to place virtual machines into physical servers of a data center. By considering the utilization of machines and node distances, our method, called Utilization Based Genetic Algorithm (UBGA), aims at reducing resource waste, network load, and energy consumption at the same time. We compared our method against several other placement methods in terms of utilization achieved, networking bandwidth consumed, and energy costs incurred, using an open-source, publicly available CloudSim simulator. The results show that our method provides better performance compared to other placement approaches.
Open Access
Application mapping algorithms for mesh-based network-on-chip architectures
(Springer New York LLC, 2015-03) Tosun, S.; Ozturk, O.; Ozkan, E.; Ozen, M.
Due to shrinking technology sizes, more and more processing elements and memory blocks are being integrated on a single die. However, traditional communication infrastructures (e.g., bus or point-to-point) cannot handle the synchronization problems of these large systems. Using network-on-chip (NoC) is a step towards solving this communication problem. Energy- and communication-efficient application mapping is a previously studied problem for mesh-based NoC architectures; however, there is still need for intelligent mapping algorithms since current algorithms either take too much running time or do not determine accurate results. To fill this need, in this study, we propose two mapping algorithms (one based on simulated annealing and one based on genetic algorithm) for energy- and communication-aware mapping problems of mesh-based NoC architectures. We compare these two algorithms with an integer linear programming-based method and a heuristic method using several multimedia and synthetic benchmarks.
Embargo
Computing artificial neural network and genetic algorithm for the feature optimization of basal salts and cytokinin-auxin for in vitro organogenesis of royal purple (cotinus coggygria scop)
(Elsevier BV, 2023-09-01) Aasim, Muhammad; Ayhan, Ayşe; Katırcı, Ramazan; Acar, Alpaslan Şevket; Ali, Seyid Amjad
This study presents the in vitro regneration protocol for Royal purple [(Cotinus coggygria Scop. (syn.: Rhus cotinus L.)] from nodal segment explants followed by optimizing the input variable combinations with the aid of PyTorch ANN and Genetic Algorithm (GA). The Murashige and Skoog (MS) culture medium yielded relatively higher regeneration frequency (91.52 %) and shoot count (1.96) as compared to woody plant medium (WPM), which yielded 84.58 % regeneration and shoot count (1.61) per explant. The supplementation of plant growth regulators (PGRs) + MS medium yielded 80.0–100.0 % shoot regeneration and 1.48–3.25 shoot counts compared to 60.0–100.0 % shoot regeneration and 1.00–2.37 shoots from the combination of PGRs + WPM. In order to predict the shoot count and regeneration with the aid of a mathematical model, the machine learning algorithms of Multilayer Perceptron (MLP), Support Vector Regression (SVR), Extreme Gradient Boosting (XGB), and Random Forest (RF) models were utilized. The highest R2 values for both output variables were acquired using MLP model in PyTorch platform. The R2 scores for regeneration and shoot counting were recorded as 0.69 and 0.71 respectively. NSGA-II algorithm revealed the 1.25 mg/L BAP (6-Benzylaminopurine), 0.02 mg/L NAA (Naphthalene acetic acid), and 0.03 mg/L IBA (Indole butyric acid) in WPM medium as an optimum combination for 100 % regeneration. On the other hand, the algorithm suggested multiple combination in MS medium for maximum shoot counting.
Open Access
Computing open-loop noncooperative solution in discrete dynamic games
(Springer, 1997) Özyildirim, S.
The purpose of this paper is to introduce a new algorithm for the approximation of non-quadratic, non-linear open-loop Nash Cournot equilibrium in a difference game of fixed duration (multiperiod) and initial state. The algorithm based on adaptive search procedure called genetic algorithm has been used to optimize strategies for N-person dynamic games. Since genetic algorithms require little knowledge of the problem itself, computations based on these algorithms are very attractive to complex dynamic optimization problems. The empirical evidences are also provided to show the success of the algorithm developed. A typical example in US macroeconomic policy selection for 1933-1936 yields evidence of political inference in the economy.
Open Access
Dual-frequency operation of probe-fed rectangular microstrip antennas with slots: analysis and design
(2001) Özgün, Özlem
Dual-frequency operation of antennas is essential for many applications in communications and radar systems, and there are various techniques to achieve this operation. Most dual-band techniques used in microstrip antennas sacrifice space, cost and weight. In this thesis, a simulation and design tool for dualband microstrip antennas, with slots on the patch and a single probe feed, is presented. This approach is based on the cavity model and modal-matching technique, where the multi-port theory is employed to analyze the effect of the slots on the input impedance. The results obtained by the simulation are verified with the experimental results. In addition, for design puiposes, a genetic algorithm is developed for the optimization of coordinates and dimension of slots in order to achieve desired frequency and impedance values for dual-frequency operation
Open Access
Fair allocation of personal protective equipment to health centers during early phases of a pandemic
(Elsevier, 2022-05) Dönmez, Zehranaz; Turhan, S.; Karsu, Özlem; Kara, Bahar Y.; Karaşan, Oya
We consider the problem of allocating personal protective equipment, namely surgical and respiratory masks, to health centers under extremely limited supply. We formulate a multi-objective multi-period non-linear resource allocation model for this problem with the objectives of minimizing the number of infected health workers, the number of infected patients and minimizing a deprivation cost function defined over shortages. We solve the resulting problem using the ε-constraint algorithm so as to obtain the exact Pareto set. We also develop a customized genetic algorithm to obtain an approximate Pareto frontier in reasonable time for larger instances. We provide a comparative analysis of the exact and heuristic methods under various scenarios and give insights on how the suggested allocations outperform the ones obtained through a set of rule-of-thumb policies, policies that are implemented owing to their simplicity and ease-of-implementation. Our comparative analysis shows that as the circumstances get worse, the trade-off between the deprivation cost and the ratio of infections deepens and that the proposed heuristic algorithm gives very close solutions to the exact Pareto frontier, especially under pessimistic scenarios. We also observed that while some rule-of-thumb policies such as a last-in-first-receives type policy work well in terms of deprivation costs in optimistic scenarios, others like split policies perform well in terms of number of infections under neutral or pessimistic settings. While favoring one of the objectives, these policies typically fail to provide good solutions in terms of the other objective; hence if such policies are to be implemented the choice would depend on the problem characteristics and the priorities of the policy makers. Overall, the solutions obtained by the proposed methods imply that more complicated distribution schemes that are not induced by these policies would be needed for best results.
Open Access
Fast and accurate mapping of complete genomics reads
(Academic Press, 2015) Lee, D.; Hormozdiari, F.; Xin, H.; Hach, F.; Mutlu, O.; Alkan C.
Many recent advances in genomics and the expectations of personalized medicine are made possible thanks to power of high throughput sequencing (HTS) in sequencing large collections of human genomes. There are tens of different sequencing technologies currently available, and each HTS platform have different strengths and biases. This diversity both makes it possible to use different technologies to correct for shortcomings; but also requires to develop different algorithms for each platform due to the differences in data types and error models. The first problem to tackle in analyzing HTS data for resequencing applications is the read mapping stage, where many tools have been developed for the most popular HTS methods, but publicly available and open source aligners are still lacking for the Complete Genomics (CG) platform. Unfortunately, Burrows-Wheeler based methods are not practical for CG data due to the gapped nature of the reads generated by this method. Here we provide a sensitive read mapper (sirFAST) for the CG technology based on the seed-and-extend paradigm that can quickly map CG reads to a reference genome. We evaluate the performance and accuracy of sirFAST using both simulated and publicly available real data sets, showing high precision and recall rates.
Open Access
Feature selection using stochastic approximation with Barzilai and Borwein non-monotone gains
(Elsevier Ltd, 2021-08) Aksakallı, V.; Yenice, Z. D.; Malekipirbazari, Milad; Kargar, Kamyar
With recent emergence of machine learning problems with massive number of features, feature selection (FS) has become an ever-increasingly important tool to mitigate the effects of the so-called curse of dimensionality. FS aims to eliminate redundant and irrelevant features for models that are faster to train, easier to understand, and less prone to overfitting. This study presents a wrapper FS method based on Simultaneous Perturbation Stochastic Approximation (SPSA) with Barzilai and Borwein (BB) non-monotone gains within a pseudo-gradient descent framework wherein performance is measured via cross-validation. We illustrate that SPSA with BB gains (SPSA-BB) provides dramatic improvements in terms of the number of iterations for convergence with minimal degradation in cross-validated error performance over the current state-of-the art approach with monotone gains (SPSA-MON). In addition, SPSA-BB requires only one internal parameter and therefore it eliminates the need for careful fine-tuning of numerous other internal parameters as in SPSA-MON or comparable meta-heuristic FS methods such as genetic algorithms (GA). Our particular implementation includes gradient averaging as well as gain smoothing for better convergence properties. We present computational experiments on various public datasets with Nearest Neighbors and Naive Bayes classifiers as wrappers. We present comparisons of SPSA-BB against full set of features, SPSA-MON, as well as seven popular meta-heuristics based FS algorithms including GA and particle swarm optimization. Our results indicate that SPSA-BB converges to a good feature set in about 50 iterations on the average regardless of the number of features (whether a dozen or more than 1000 features) and its performance is quite competitive. SPSA-BB can be considered extremely fast for a wrapper method and therefore it stands as a high-performing new feature selection method that is also computationally feasible in practice.
Open Access
Genetic algorithm for closed-loop equilibrium of high-order linear-quadratic dynamic games
(Elsevier, 2000) Özyıldırım, S.
In this paper, we implement an adaptive search algorithm, genetic algorithm to derive closed-loop Nash equilibria for linear-quadratic dynamic games. The computation of these equilibria is quite difficult to deal with analytically and numerically. Our strategy is to search over all time-invariant strategies depending only on the current value of the state. Also provided are some evidences which show the success of the algorithm.
Open Access
Heterogeneous network-on-chip design through evolutionary computing
(Taylor & Francis, 2010) Ozturk, O.; Demirbas, D.
This article explores the use of biologically inspired evolutionary computational techniques for designing and optimising heterogeneous network-on-chip (NoC) architectures, where the nodes of the NoC-based chip multiprocessor exhibit different properties such as performance, energy, temperature, area and communication bandwidth. Focusing primarily on array-dominated applications and heterogeneous execution environments, the proposed approach tries to optimise the distribution of the nodes for a given NoC area under the constraints present in the environment. This article is the first one, to our knowledge, that explores the possibility of employing evolutionary computational techniques for optimally placing the heterogeneous nodes in an NoC. We also compare our approach with an optimal integer linear programming (ILP) approach using a commercial ILP tool. The results collected so far are very encouraging and indicate that the proposed approach generates close results to the ILP-based approach with minimal execution latencies. © 2010 Taylor & Francis.
Open Access
Hybrid model for probe-fed rectangular microstrip antennas with shorting strips
(2001) Mutlu, Selma
In the dual frequency operation of microstrip antennas, shorting strips are used to adjust the ratio of frequencies. A multi-port analysis is usually employed to predict the input impedance and resonant frequency of probe-fed microstrip antennas with shorting strips. However, this approach does not provide any information about the field distribution under the patch. In this thesis, a hybrid model, using both the cavity model and point matching, is developed to calculate the field distribution under the patch with shorting pins and strips. In addition, this model also accounts for the conducting nature of the feed and shorting strips, with the help of the point-matching algorithm. Then, to verify the model, the theoretical results obtained from the hybrid method are compared to the experimental results and good agreement is observed. Finally, a genetic algorithm is developed for optimizing the position and width of the shorting strips to achieve desired frequency ratio and input impedances in dual-band operations.
Open Access
A problem space genetic algorithm in multiobjective optimization
(Springer New York LLC, 2003) Türkcan, A.; Aktürk, M. S.
In this study, a problem space genetic algorithm (PSGA) is used to solve bicriteria tool management and scheduling problems simultaneously in flexible manufacturing systems. The PSGA is used to generate approximately efficient solutions minimizing both the manufacturing cost and total weighted tardiness. This is the first implementation of PSGA to solve a multiobjective optimization problem (MOP). In multiobjective search, the key issues are guiding the search towards the global Pareto-optimal set and maintaining diversity. A new fitness assignment method, which is used in PSGA, is proposed to find a well-diversified, uniformly distributed set of solutions that are close to the global Pareto set. The proposed fitness assignment method is a combination of a nondominated sorting based method which is most commonly used in multiobjective optimization literature and aggregation of objectives method which is popular in the operations research literature. The quality of the Pareto-optimal set is evaluated by using the performance measures developed for multiobjective optimization problems.
Open Access
A resampling-based meta-analysis for detection of differential gene expression in breast cancer
(BioMed Central, 2008) Gur-Dedeoglu, B.; Konu, O.; Kir, S.; Ozturk, A. R.; Bozkurt, B.; Ergul, G.; Yulug, I.G.
Background: Accuracy in the diagnosis of breast cancer and classification of cancer subtypes has improved over the years with the development of well-established immunohistopathological criteria. More recently, diagnostic gene-sets at the mRNA expression level have been tested as better predictors of disease state. However, breast cancer is heterogeneous in nature; thus extraction of differentially expressed gene-sets that stably distinguish normal tissue from various pathologies poses challenges. Meta-analysis of high-throughput expression data using a collection of statistical methodologies leads to the identification of robust tumor gene expression signatures. Methods: A resampling-based meta-analysis strategy, which involves the use of resampling and application of distribution statistics in combination to assess the degree of significance in differential expression between sample classes, was developed. Two independent microarray datasets that contain normal breast, invasive ductal carcinoma (IDC), and invasive lobular carcinoma (ILC) samples were used for the meta-analysis. Expression of the genes, selected from the gene list for classification of normal breast samples and breast tumors encompassing both the ILC and IDC subtypes were tested on 10 independent primary IDC samples and matched non-tumor controls by real-time qRT-PCR. Other existing breast cancer microarray datasets were used in support of the resampling-based meta-analysis. Results: The two independent microarray studies were found to be comparable, although differing in their experimental methodologies (Pearson correlation coefficient, R = 0.9389 and R = 0.8465 for ductal and lobular samples, respectively). The resampling-based meta-analysis has led to the identification of a highly stable set of genes for classification of normal breast samples and breast tumors encompassing both the ILC and IDC subtypes. The expression results of the selected genes obtained through real-time qRT-PCR supported the meta-analysis results. Conclusion: The proposed meta-analysis approach has the ability to detect a set of differentially expressed genes with the least amount of within-group variability, thus providing highly stable gene lists for class prediction. Increased statistical power and stringent filtering criteria used in the present study also make identification of novel candidate genes possible and may provide further insight to improve our understanding of breast cancer development.
Open Access
Three-country trade relations: a discrete dynamic game approach
(Elsevier, 1996) Özyıldırım, S.
A three-country, two-bloc trade model is used to determine the impact of a coalition within the blocs on the optimal pricing policies of the bloc. It is shown in a North-South world where the South has to cooperate for efficient pricing policy. In addition to the complexities of interactions between three countries, a dynamic game approach leads to the usage of numerical methods in this paper. We used a new algorithm based on adaptive search procedure called genetic algorithm to optimize strategies for three-person discrete dynamic games. Welfare implications are also addressed.
Open Access
A utilization based genetic algorithm for virtual machine placement in cloud computing systems
(2016-09) Çavdar, Mustafa Can
Due to increasing demand for cloud computing and related services, cloud providers need to come up with methods and mechanisms that increase performance, availability and reliability of datacenters and cloud computing systems. Server virtualization is a key component to achieve this, which enables sharing of resources of a physical machine among multiple virtual machines in a totally isolated manner. Optimizing virtualization has a very signi cant e ect on the overall performance of cloud computing systems. This requires e cient and effective placement of virtual machines into physical machines. Since this is an optimization problem that involves multiple constraints and objectives, we propose a method based on genetic algorithms to place virtual machines. By considering utilization of machines and node distances, our method aims at reducing resource waste, network load, and energy consumption at the same time. We compared our method with several other methods in terms of utilization achieved, networking bandwidth consumed, and energy costs incurred, using the publicly available CloudSim simulation platform. The results show that our approach provides improved performance compared to other similar approaches.