Bildiri Koleksiyonu

Bu koleksiyon için kalıcı URI

https://hdl.handle.net/20.500.12684/6862

Listeleniyor 1 - 20 / 54

Kablosuz Algılayıcı Ağlarda Kullanılan Enerji Duyarlı Kümeleme Algoritmaları
(Ieee, 2019) Tay, Muhammet; Şentürk, Arafat
Kablosuz Algılayıcı Ağ (KAA)'lar kullanıldıkları alanlar gereği enerjiye her zaman ihtiyaç duymaktadırlar. Bu sebeple var olan enerjilerini en verimli şekilde kullanmaları gerekmektedir. KAA'lar için enerji verimliliği sağlanmasında en önemli rollerden birisi algılayıcı düğümler arasında kümeleme oluşturmaktır. Kümelenen algılayıcılar içinde belirlenen kriterlere göre en uygun algılayıcı düğümü Küme Başı (KB) olarak seçmek enerji sarfiyatını azaltmaktadır. Bu çalışmada, literatürde en çok bahsedilen kümeleme algoritmalarına yer verilmiş ve bu algoritmalar belirli ölçütler çerçevesinde karşılaştırılmıştır. Karşılaştırma sonucunda kümeleme algoritmalarının avantaj ve dezavantajları belirtilmiştir.
Geniş Ölçekli Kablosuz Algılayıcı Ağlar İçin Enerji Verimli Melez OEK Protokolü
(Ieee, 2015) Karahan, Alper; Ertürk, İsmail; Atmaca, Sedat; Çakıcı, Süleyman
Bu bildiride sunulan çalışmada geniş ölçekli KAA (Kablosuz Algılayıcı Ağ) uygulamalarında kullanılmak üzere düğüm enerji tüketim verimliliğini esas alan melez bir OEK (Ortam Erişim Kontrol) protokolü geliştirilmiştir. Tasarlanan melez OEK protokolünde, CSMA (Carrier Sense Multiple Access) ve TDMA (Time Division Multiple Access) ortam erişim yöntemlerinin üstün yönleri bir arada kullanılmaktadır. Böylece geniş ölçekli KAA uygulamalarında her iki yöntemden de daha enerji verimli bir OEK protokolü hedeflenmiştir. Olaylara (events) hızlı tepki veren CSMA yöntemi ile çarpışma ve gereksiz dinlemeleri en aza indiren TDMA yöntemi alma tabanlı olarak bütünleştirilerek yüksek sistem başarımı ve düğüm enerji verimliliği elde edilmiştir.
Hybrid FCM-WOA Data Clustering Algorithm
(Ieee, 2018) Arslan, Hatice; Toz, Metin
In this work, we propose a hybrid clustering algorithm that integrates Fuzzy C-Means (FCM) and Whale Optimization Algorithm (WOA) using the Chebshev distance function. The FCM algorithm uses Euclidean distance to measure the similarity between the data. To avoid the existing disadvantages of the Euclidean distance, all distances in the FCM algorithm is calculated with the Chebsyhev distance function. The BOA algorithm is used to optimize the initial cluster centers. The proposed hybrid algorithm is tested with three different sets of data selected from UCI Machine Learning Repository database. As a result, it is seen that the clustering performance of the proposed algorithm is much better than the FCM algorithm.
g-BSAFCM : A New Hybrid Clustering Algorithm
(Ieee, 2016) Toz, Güliz; Erdoğmuş, Pakize
Clustering is dividing a dataset into subsets that has similar characteristics. In this study, fuzzy c-means clustering algorithm (FCM) and a new evolutionary optimization algorithm, Backtracking Search (BSA) algorithm, were combined and a new hybrid clustering algorithm (BSAFCM) was proposed. Moreover, the local search abilities of the new algorithm was improved and the new algorithm was named as g-BSAFCM. Three benchmark datasets from UCI Machine Learning Repository database were clustered by using the developed algorithms and FCM. According to the results g-BSAFCM has achieved better results than FCM and BSAFCM.
Fuzzy Logic Based Channel Selection for Mobile Secondary Users in Cognitive Radio Networks
(Ieee, 2015) Bayrakdar, Muhammed Enes; Çalhan, Ali
Cognitive radio is a new technology that aims to solve the spectrum scarcity problem. In cognitive radio networks; because secondary users utilize the spectrum in an opportunistic manner, it is crucial not to cause any interference to the primary users. In this work, a new approach for channel assignment to mobile secondary users by means of fuzzy logic is proposed. In the proposed fuzzy logic system; by using input parameters of density, mobility, and noise, channel selection probability is obtained. Density parameter expresses how intense the spectrum is used by primary users. Mobility parameter is a distance among secondary users, primary users, and base station. Noise parameter represents the noise, and other disruptive effects in the spectrum. Channel selection probability shows which frequency band of primary users to use according to the parameters of mobile secondary user. In the proposed system, the most suitable channel selection is provided for mobile secondary users.
Fuzzy Logic Based Spectrum Handoff Decision for Prioritized Secondary Users in Cognitive Radio Networks
(Ieee, 2015) Bayrakdar, Muhammed Enes; Çalhan, Ali
Recent studies have revealed that due to the increasing number of wireless users, spectrum scarcity problem has arisen. In order to alleviate this problem, cognitive radio technology has emerged with the aim of using available spectrum in an opportunistic manner by secondary users. While utilizing licensed spectrum, it is inevitable for secondary users not to cause any interference to the primary users. Once the primary user activity is detected on the licensed channel at the time of on-going transmission of secondary user, secondary user must either change the licensed spectrum or stop its transmission. This case of changing spectrum by secondary user is known as spectrum handoff. In this paper, fuzzy logic based spectrum handoff decision is carried out by using data rate, channel usage, and priority. Different data traffics such as audio, and video are considered for secondary users. In order to simulate data traffics, two different simulation scenarios are realized. Besides, several priority classes are taken into account for urgent and real-time communications. In the proposed system, it is seen that accurate spectrum handoff decisions are made for different traffic types.
Font and Turkish Letter Recognition in Images with Deep Learning
(Ieee, 2018) Sevik, Aylin; Erdoğmuş, Pakize; Yalçın, Erdi
The purpose of this article is to recognize letter and especially font from images which are containing texts. In order to perform recognition process, primarily, the text in the image is divided into letters. Then, each letter is sended to the recognition system. Results are filtered according to vowels which are most used in Turkish texts. As a result, font of the text is obtained. In order to separate letters from text, an algorithm used which developed by us to do separation. This algorithm has been developed considering Turkish characters which has dots or accent such as i, j, u, o and g and helps these characters to be perceived by the system as a whole. In order to provide recognition of Turkish characters, all possibilities were created for each of these characters and the algorithm was formed accordingly. After recognizing the each character, these individual parts are sended to the pre-trained deep convolutional neural network. In addition, a data set has been created for this pre-trained network. The data set contains nearly 13 thousands of letters with 227(star)227(star)3 size have been created with different points, fonts and letters. As a result, 100 percent of success has been attained in the training. %79.08 letter and %75 of font success has been attained in the tests.
Determination of Type and Quality of Hazelnut using Image Processing Techniques
(Ieee, 2015) Bayrakdar, Sümeyye; Çomak, Bekir; Başol, Derya; Yücedağ, İbrahim
Hazelnut (Corylus avellana L.) comes number two after almond in the ranking of hard shelled fruits that is cultivated commonly in the world. According to the FAO (Food and Agriculture Organization) statistics, Turkey covers approximately 70% of the world hazelnut production, and 82% of the hazelnut exportation. These statistics indicate that Turkey is the first largest hazelnut producer and exporter in the world. Quality and automation have a great importance in the agricultural industry that put our country forward as world-wide leader. In the quality control systems, classification studies with image processing methods have accelerated in recent years. In this study, it is aimed that determination of type and quality of shelled hazelnuts with image processing by using size and shape characteristics of hazelnuts. As a result of studies on a variety of hazelnuts, it was reached 84% accuracy rate for grouping of shelled hazelnut according to the type and commercial definitions. The quality of hazelnuts also was acquired without error. Inshell Hazelnut Standard (TS 3074) that is published by the Turkish Standards Institute (TSE) is based for shelled hazelnuts.
Data Clustering Based on FCM and WOA
(Ieee, 2018) Arslan, Hatice; Toz, Metin
In this article, we have proposed a new hybrid clustering algorithm based on Fuzzy C Means (FCM) and Whale Optimization Algorithm (WOA) using chaotic map. The random selection of the initial cluster centers in the FCM algorithm is a disadvantage for the algorithm. To reduce this disadvantage, the BOA algorithm, which is improved performance with chaotic maps, is used to optimize the initial cluster centers. The proposed hybrid algorithm is tested for four different sets of data and better results are obtained from the FCM algorithm.
Delay Characteristics of TDMA Medium Access Control Protocol for Cognitive Radio Networks
(Ieee, 2016) Bayrakdar, Muhammed Enes; Çalhan, Ali
In this work, we have evaluated the delay characteristics of Time Division Multiple Access (TDMA) protocol. In our simulation scenarios, primary users and secondary users exploit TDMA as a medium access control protocol. We have designed a network environment in Riverbed (OPNET) simulation software that consists of primary users, secondary users, and base stations. In our network model, secondary users sense the spectrum and inform the base station about empty channels. Then, the base station decides accordingly which secondary user may exploit the empty channel. Energy detection technique is employed as a spectrum sensing technique because it is the best when information about signal of primary user is obtained. Besides, different number of users is selected in simulation scenarios in order to achieve accurate delay results. Comparing analytical model with simulation results, we have shown that delay analysis of our system model is reliable and correct.
Cuff-Less Continuous Blood Pressure Estimation from Electrocardiogram(ECG) and Photoplethysmography (PPG) Signals with Artificial Neural Network
(Ieee, 2018) Şentürk, Ümit; Yücedağ, İbrahim; Polat, Kemal
Continuous blood measurement important information about the health status of the individuals. Conventional methods use a cuff for blood pressure measurement and cannot be measured continuously. In this study, we proposed a system that estimates systolic blood pressure (SP) and diastolic blood pressure (DP) for each heart beat by extracting attributes from ECG and PPG signals. Simultaneous ECG and PPG signals from the PhysioNet Database are pre-processed (denoising, artifact cleaning and baseline wandering) to remove noise and artifacts and segmented into R-R peaks. For each heartbeat, 22-time domain features were extracted from ECG and PPG signals. SP and DP values were estimated by introducing these 22 attributes to the model of Lavenberg-Marquardt artificial neural networks (ANN). Arterial blood pressure (ABP) was also taken from the PhysioNet MIMIC II database along with ECG and PPG signals. ABP signals have been used as targets in the artificial neural network. The system performance has been evaluated by calculating the difference between the estimated ABP values and the actual by the ANN model. The performance value between the predicted SP and actual SP values is -0.14 +/- 2.55 (mean +/- standard deviation) and the performance value between estimated DP and actual DP values is -0.004 +/- 1.6. The obtained results have shown that the proposed model has predicted blood pressure with high accuracy. In this study, SP and DP values can also be measured directly without any calibration in blood pressure estimation.
TSCBAS: A Novel Correlation Based Attribute Selection Method and Application on Telecommunications Churn Analysis
(Ieee, 2018) Kayaalp, Fatih; Başarslan, Muhammet Sinan; Polat, Kemal
Attribute selection has a significant effect on the performance of the machine learning studies by selecting the attributes having significant effect on result, reducing the number of attributes, and reducing the calculation cost. In this study, a new attribute selection method which is a combination of the Rcorrelation coefficient-based attribute selection (RCBAS) and the rho-correlation coefficient-based attribute selection (rho CBAS) called the Two-Stage Correlation-Based Attribute Selection (TSCBAS) is proposed to select significant attributes. The proposed attribute selection method has been applied to customer churn prediction on a telecommunications dataset for performance evaluation. The dataset used in the study includes real customer call records details for the years 2013 and 2014 obtained from a major telecommunications company in Turkey. Apart from the proposed attribute selection method, four different methods named Rcorrelation coefficient-based attribute selection, rho-correlation coefficient-based attribute selection, ReliefF, and Gain Ratio have been used for creating five datasets. After that, four classifier algorithms including Random Forest, C4.5 Decision Tree, Naive Bayes and AdaBoost. M1 have been applied. The obtained results have been compared according to the performance metrics comprising Accuracy (ACC), Sensitivity (TPR), Specificity (SPC), F-measure (F), AUC (area under the ROC curve), and run-time. The results of the comparisons show that the proposed attribute selection algorithm outperforms the state of the art methods on customer churn prediction.
The Future of Wireless Technology and Potential Problems and Solutions
(Ieee, 2017) Yılmaz, Şeyhmus; Toklu, Sinan
Wireless is among technology's biggest contributions to human beings. Wireless technology involves the delivering of data over a distance without aid of wires, cables or any other types of electrical conductors. The delivering distance could be anyplace among a few meters and thousands of kilometers. In this paper, we outline the future of wireless technology and discuss potential problems and solutions.
The Classification of Breast Cancer with Machine Learning Techniques
(Ieee, 2016) Kolay, Nurdan; Erdoğmuş, Pakize
In this study, it is aimed to classify breast cancer data attained from UCI(University of California-Irvine), Machine Learning Laboratory with some Machine Learning Techniques. With this aim, clustering performance of some distance measures in Matlab(C) has been compared, using breast cancer data. Later without using any pre-processing, some of the machine learning techniques are used for the clustering breast cancer data, using WEKA data mining software(C). As a result, it has been seen that distance measures effects the clustering performance nearly 12 percentage and the succes of the classification varies from % 45 to % 79, according to the methods.
Stock Market Prediction with Deep Learning Using Financial News
(Ieee, 2018) Gündüz, Hakan; Yaslan, Yusuf; Çataltepe, Zehra
In this study, the hourly movement directions of 9 banking stocks in Borsa Istanbul were predicted using Long-Short Term Memory(LSTM) networks with features obtained from financial news. In the feature creation phase, the word embedding referred as Fasttext, and the financial sentiment dictionary were utilized. Class labels indicating the movement direction were computed based on hourly close prices of the stocks and they were aligned with obtained feature vectors. Two different LSTM networks were trained to perform the prediction, and the performance of the classification process was evaluated by the Macro Averaged (M.A) F-Measure. In the experiments, the movement directions of the 9 stocks were predicted with an average M.A F-measure rate of 0.540. Although the results of both LSTM networks were higher than the Random and Naive benchmark methods, the use of Attention Mechanism in the second LSTM network did not positively affect the results.
Spectrum Handoff Process with Aging Solution for Secondary Users in Priority based Cognitive Networks
(Ieee, 2017) Bayrakdar, Muhammed Enes; Çalhan, Ali
Spectrum handoff is a process performed by secondary users in cognitive radio networks. To accomplish this, the base station that coordinates the secondary users decides which user will make the spectrum handoff process according to certain criteria. The priority classes of the secondary users are at the top of these criteria. In traditional priority based queues, packets with higher priority are transmitted first, and lower priority packets waits for the other higher-priority packet transmissions to finish. In this study, priority data traffic is used in queue structure in order to meet the different requirements of secondary users. In addition, we have added the aging solution to the spectrum handoff mechanism in order to shorten the long wait times of low priority packets. The aging solution is defined as the increase of the priority of lower priority packets waiting for too long in the queue over certain waiting periods. Analytical and simulation models of the aging solution have been designed in order to prove validation. It has been shown that the total number of spectrum handoff on the network for different priority packets and different loads is reduced significantly.
Snow Flake Optimization Algorithm
(Ieee, 2018) Akkoyun, Özgün; Toz, Metin
Optimization is the best result finding process under certain constraints for a given problem. In this study, a new population-based optimization algorithm, Snow Flake Optimization Algorithm (SFO), is proposed to model a snowflake falling on the ground. The mathematical formulation of the developed algorithm is given. In addition, 8 benchmark functions are solved with the proposed algorithm and the results are presented in tabular form. According to the results, SFO achieved successful results in solving all functions.
Simulation Model of Spectrum Handoff Process for Medium Access Control Protocols in Cognitive Radio Networks
(Ieee, 2015) Bayrakdar, Muhammed Enes; Çalhan, Ali
Cognitive radio is a new technology that improves spectrum utilization by allowing secondary users to use licensed channels in an opportunistic manner. Because primary users have a right to use their licensed channels at any time, it is necessary for secondary users to stop their transmissions or to continue their transmissions in other channels in case of action of primary users. The event of changing channel for secondary users is known as spectrum handoff. Spectrum handoff for medium access protocols is an emerging topic because of the increasing spectrum scarcity. In this paper, simulation model of a cognitive radio network that consists of primary users and secondary users is considered. Riverbed Modeler Simulation Software is used for simulation scenarios. Secondary users are designed to make spectrum handoff between time slots of primary users with reactive decision spectrum handoff. It is clearly seen that number of handoff is significantly decreased by means of instant sensing based reactive spectrum handoff.
SDN-Enabled Wireless Body Area Networks
(Ieee, 2018) Cicioğlu, Murtaza; Çalhan, Ali
Network management processes of Wireless Body Area Networks (WBANs) such as installation and configuration are quite complex because of heterogeneous structure and limited resources of WBANs. In addition, the lack of a manageable and flexible structure poses an important problem in WBANs. The software-defined network (SDN) approach suggests a new network architecture that is simple, flexible, and manageable and has less workload. This approach is considered to be a solution to the above-mentioned problems of the WBAN architecture. In this context, WBAN architecture based on SDN approach, a new network approach for WBANs, is proposed in this paper. A controller as SDN control unit is responsible for all network-related management and control operations in WBANs. This unit effectively and efficiently manages all the wireless communication processes necessary for the coordinator nodes to communicate with each other and with the controller. The throughput and end-to-end delay results of the proposed architecture are examined for the performance analysis. The results show that the proposed network architecture improves the performance of the traditional WBANs structure and simplifies the control and management processes.
Repetitive neural network (RNN) based blood pressure estimation using PPG and ECG signals
(Ieee, 2018) Şentürk, Ümit; Yücedağ, İbrahim; Polat, Kemal
In this study, a new hybrid prediction model was proposed by combining ECG (Electrocardiography) and PPG (Photoplethysmographic) signals with a repetitive neural network (RNN) structure to estimate blood pressure continuously. The proposed method consists of two steps. In the first step, a total of 22 time-domain attributes were obtained from PPG and ECG signals to estimate systolic blood pressure (SBP) and diastolic blood pressure (DBP) values. In the second phase, these time-domain attributes are set as input to the RNN model and then the blood pressure is estimated. Within the RNN structure, there are two-way long short-term memory BLSTM (Bidirectational Long-Short Term Memory), LSTM and ReLU (Rectified-Linear unit) layers. The bidirectional LSTM layer has been used to remove the negative affects the blood pressure value of past and future effects of nonlinear physiological changes. The LSTM layers has ensured that learning is deep and that mistakes made are reduced. The ReLU layer has been allowed the neural network to quickly reach its conclusion. The same ECG and PPG signals obtained from the database have been cleared from noise and artifacts. And then ECG and PPG signals have been segmented according to peak values of these signals. The results have shown that RMSE (Root Mean Square Error) between the estimated SBP and the measured SBP with RNN model was 3.63 and the RMSE between the estimated DBP and the measured DBP values was 1.48 with RNN model. It has been seen that the used model has a more learning ability. Thanks to the proposed method, a calibration free blood pressure measurement system using PPG and ECG signals, was developed.

Güncel Gönderiler