Heart disease is one of the most known and deadly diseases in the world, and many people lose the... more Heart disease is one of the most known and deadly diseases in the world, and many people lose their lives from this disease every year. Early detection of this disease is vital to save people’s lives. Machine Learning (ML), an artificial intelligence technology, is one of the most convenient, fastest, and low-cost ways to detect disease. In this study, we aim to obtain an ML model that can predict heart disease with the highest possible performance using the Cleveland heart disease dataset. The features in the dataset used to train the model and the selection of the ML algorithm have a significant impact on the performance of the model. To avoid overfitting (due to the curse of dimensionality) due to the large number of features in the Cleveland dataset, the dataset was reduced to a lower dimensional subspace using the Jellyfish optimization algorithm. The Jellyfish algorithm has a high convergence speed and is flexible to find the best features. The models obtained by training the ...
Supervisory Control and Data Acquisition (SCADA) systems play a crucial role in overseeing and co... more Supervisory Control and Data Acquisition (SCADA) systems play a crucial role in overseeing and controlling renewable energy sources like solar, wind, hydro, and geothermal resources. Nevertheless, with the expansion of conventional SCADA network infrastructures, there arise significant challenges in managing and scaling due to increased size, complexity, and device diversity. Using Software Defined Networking (SDN) technology in traditional SCADA network infrastructure offers management, scaling and flexibility benefits. However, as the integration of SDN-based SCADA systems with modern technologies such as the Internet of Things, cloud computing, and big data analytics increases, cybersecurity becomes a major concern for these systems. Therefore, cyber-physical energy systems (CPES) should be considered together with all energy systems. One of the most dangerous types of cyber-attacks against SDN-based SCADA systems is Distributed Denial of Service (DDoS) attacks. DDoS attacks disrupt the management of energy resources, causing service interruptions and increasing operational costs. Therefore, the first step to protect against DDoS attacks in SDN-based SCADA systems is to develop an effective intrusion detection system. This paper proposes a Decision Tree-based Ensemble Learning technique to detect DDoS attacks in SDN-based SCADA systems by accurately distinguishing between normal and DDoS attack traffic. For training and testing the ensemble learning models, normal and DDoS attack traffic data are obtained over a specific simulated experimental network topology. Techniques based on feature selection and hyperparameter tuning are used to optimize the performance of the decision tree ensemble models. Experimental results show that feature selection, combination of different decision tree ensemble models, and hyperparameter tuning can lead to a more accurate machine learning model with better performance detecting DDoS attacks against SDN-based SCADA systems.
Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji, 2018
Günümüzde giderek artan bir oranda makinelerin birbirleri ile iletişimde olduğu görülmektedir. He... more Günümüzde giderek artan bir oranda makinelerin birbirleri ile iletişimde olduğu görülmektedir. Herhangi bir sektör için geliştirilen makineler arası iletişim (M2M: Machine to Machine) uygulamasının kullanıcı ile etkileşimi için bir M2M platformuna ihtiyaç duyulmaktadır. M2M platformunun amacı veri toplama, veri izleme, veri yönetimi ve raporlama-analiz gibi işlevleri sağlamaktır. M2M platformları çoğunlukla web tabanlı olarak geliştirilir ve kullanıcının her ortamdan platforma erişim sağlaması hedeflenir. Bu hedef doğrultusunda bazı M2M platformları hem masaüstü hem de mobil uygulama sürümleri oluşturularak kullanıcıya sunulmaktadır. Fakat bu durum mobil işletim sistemi ve mobil ekran boyutuna göre farklı arayüz tasarımları yapmayı gerektirmektedir. Bu durum arayüz geliştiricilerinin işini oldukça zorlaştırmaktadır. Bu çalışmada her mobil işletim sistemi ve ekran boyutuna göre arayüz geliştirmek yerine duyarlı tasarıma sahip web tabanlı bir M2M platformu geliştirilmiştir. Bu sayede ...
Word error rate is high when the acoustic model used in Turkish Automatic Speech Recognition (ASR... more Word error rate is high when the acoustic model used in Turkish Automatic Speech Recognition (ASR) systems is not supported by a robust language model. The use of a welldesigned language model in combination with an acoustic model in the ASR reduces the word error rate. In the training of the language model required for ASR, plain text data is used. The accuracy of the text data used is very important for the training of ASR models. In this study, the detection and correction of spelling errors in the text data used in the education of Turkish ASR system were realized by using a method based on natural language processing. Firstly, misspelled words were found in the text data. Lack of characters, surplus characters,
Türkiye Bilişim Vakfı Bilgisayar Bilimleri ve Mühendisliği Dergisi, Jul 6, 2017
Teknolojinin ve internet alt yapısının hızlı gelişimi, gömülü cihazları kullanarak insanların ile... more Teknolojinin ve internet alt yapısının hızlı gelişimi, gömülü cihazları kullanarak insanların iletişim kurma şeklini değiştirmiştir. Mevcut durumda fiziksel nesnelerin büyük bir kısmının iletişim yeteneği bulunmamaktadır. Nesnelerin interneti (IoT: Internet of Things) yaklaşımı ve Makinalar Arası İletişim (M2M: Machine to Machine) teknolojileri sayesinde nesnelere iletişim yeteneği kazandırılabilmektedir. Bu sayede nesnelerin uzaktan izlenmesi ve kontrolü gerçekleştirilmektedir. M2M/IoT uygulamalarının yönetilebilmesi için M2M/IoT platformları kullanılmaktadır. Bu çalışmada, öncelikle M2M/IoT teknolojisi açıklanmış M2M/IoT platformları üzerinde prototip uygulama geliştirilmiştir. Kullanılan platformlar mevcutta ticari olarak hizmet vermektedir. Bu platformların genel yapısı ve mimarisi, hangi alanlarda kullanıldığı, platform üzerinde uygulamaların nasıl geliştirildiği incelenmiştir. Uygulamaya bağlı olarak M2M/IoT platformlarının farklı işlevsel özelliklere sahip oldukları görülmüştür.
Machine to machine (M2M) is a smart ecosystem that enables devices to communicate by exchanging d... more Machine to machine (M2M) is a smart ecosystem that enables devices to communicate by exchanging data through specific protocols autonomously with limited human interaction (or no human interaction) between each other. Use cases, business models, varieties of devices and protocols of M2M systems getting proliferating. Being able to accommodate this large spectrum of possibilities requires a standardized M2M platform design based on abstraction, modularity, and scalability. In this study, a web-based scalable M2M platform was developed which is compatible with European Telecommunications Standards Institute (ETSI) and oneM2M standards. The web-based M2M service platform has with a sensitive design adopting Web 3.0 standards and Service Oriented Architecture (SOA) approaches. The access to M2M services and data was provided using Representational State Transfer, RestFul Web Services (REST) through a service layer. The data acquired from M2M devices were stored in a non-relational database. The security of M2M platform was achieved through token based ID check. Fast, scalable, cost-efficient, secure, standardized, general purpose an M2M service platform was created.
Uluslararası bilimsel araştırmalar dergisi, Aug 1, 2019
Respiratory sounds are perceived as the most important diagnostic value by many physicians in dia... more Respiratory sounds are perceived as the most important diagnostic value by many physicians in diagnosing patients suffering from lung disease and monitoring their clinical course. It is no doubt that the results of visual-acoustic analysis of lung sounds in the computer will contribute to the diagnosis of lung diseases. Acoustic analysis of respiration can quantify changes in lung sounds, make permanent recordings of lung sounds and produce graphic representations. Thus, monitoring and diagnosis of patients suffering from common lung diseases can be facilitated. In this study, time amplitude waveform, time expanded waveform, RMS values, frequency changes cohesive to zero crossing velocity were observed in the time domain. At the same time, thanks to the algorithms developed in the time domain, the wheeze and crackle sounds from the lung sounds were determined automatically. According to pre-defined amplitude and time criteria when an unusual event that represents wheeze and crackle sounds in various parts of the lung sounds, the detection of unusual event was performed. As a result, time-domain analysis of lung sounds has been shown to provide more information on the diagnosis of lung diseases in a non-invasive method.
The objective of this study is to investigate the effects of machining parameters on the acoustic... more The objective of this study is to investigate the effects of machining parameters on the acoustic frequency developed during the machining of cold work tool steels. The machining tests were carried out through a milling method with different machining parameters. Analogue acoustic signals obtained via the microphone were converted and digitized. During idle operation and cutting operation, the obtained spindle sound recordings were subjected to Fast Fourier Transform (FFT) and converted from the time domain to the frequency domain, and the statistical effects of cutting parameters were investigated by use of analysis of variance. Cutting speed was found to be the only influential factor for the acoustic frequency in idle time. In the starting of machining, the feed rate, cutting speed, and depth of cut were seen to affect the acoustic frequency, while the cutting speed, insert material, insert radius, depth of cut, and feed rate were seen to be effective in the later stages of machining.
In this study, developed a platform that performs content analysis on the YouTube videos. With th... more In this study, developed a platform that performs content analysis on the YouTube videos. With the developed platform, it is aimed to identify the videos supporting the terrorist organizations on YouTube and to remove the detected videos. In the study, an image analysis model was constructed based on the Convolutional Neural Network (CNN) and the AlexNet model. Were made separately analysis performed on each of the frame forming the video image. For the analysis of the images contained in the video, training data set is prepared from different YouTube videos. The analysis platform was developed with a RESTful web service based architecture using Spring Framework. A web interface for users to interact with the platform has been developed. Each user can receive analysis results of different videos via email notification. For platform testing, videos with and without terrorist propaganda were used. In the tests conducted after CNN network training, 70% accuracy rate was achieved. By increasing the number of content in the training data set of video analysis process it is thought to increase the success rate.
Bu calismada lazer ve cizgisel CCD(Charge-Coupled Device) algilayici tabanli bir optik cap olcum ... more Bu calismada lazer ve cizgisel CCD(Charge-Coupled Device) algilayici tabanli bir optik cap olcum sistemi gelistirilmistir. Bu sekilde bir olcum sistemi ile optik temassiz cap olcumleri gerceklestirilmistir. Bir lazer diyot ve kollimator lens kullanarak paralel bir isik demeti olusturulmustur. Paralel isik demeti karsisina cizgisel CCD algilayici yerlestirilmis ve lazer isik demeti ile aydinlatilmistir. Paralel isik demeti icerisine bir nesne konumlandirilip cizgisel CCD algilayici uzerine nesne golgesi dusurulmustur. Bu sekilde cizgisel CCD algilayicida golge altinda kalan kisimlardaki pikseller aydinlanmadigindan, nesnenin capi olusan golgeden belirlenebilmektedir. Calismada silindirik kesit alanli nesnelerin capinin temassiz olarak olculmesi %0,24 - %1,34 hata payi ile gerceklestirilmistir.
Akciğer hastalıklarının tanısında yeni enstrümanlar ve akciğer ses analizleri klinik uygulamalard... more Akciğer hastalıklarının tanısında yeni enstrümanlar ve akciğer ses analizleri klinik uygulamalarda hala yaygın olarak kullanılmamaktadır. Bu çalışmada solunum problemine sahip 48 hastanın oskültasyonu deneyimli bir doktor tarafından elektronik stetoskop (Littmann Electronic Stethoscope Model 4000) kullanılarak yapıldıktan sonra akciğer sesleri stetoskoba kaydedilmiş ve bu sesler doktorun hastaya ait yorumları ile beraber hastanın sürekli takibi amacıyla bilgisayara bir hasta veritabanı olarak kaydedilmiştir. Ayrıca stetoskopta bulunan sesler bilgisayara uygun ses formatında da kaydedilerek daha sonra akustik analizi için olanak sağlanmıştır. Böylece hem gerçek zamanlı dinleme hem de bilgisayar akustik analiz metotlarının kullanılması ile akciğer seslerindeki değişimler, doktorlara normal ve anormal solunum sesi bilgisini sağlamış ve daha kolay teşhis imkanı oluşturmuştur.
In response to the similar challenges in building large-scale distributed applications and platfo... more In response to the similar challenges in building large-scale distributed applications and platforms on the Web, microservice architecture has emerged and gained a lot of popularity in recent years. Therefore, both for the use of microservices and for the provided of the necessary interface for Automatic Speech Recognition (ASR), a web-based platform has been developed. Within firstly the scope of the study, a Turkish ASR system was developed. A web service structure was created to facilitate access to the ASR system. The access of methods and data in the web service structure was provided through Representational State Transfer (REST) web services and service layer. An interface was developed to enable interaction with the web service. The platform was developed using a combination of different technologies such as ASR, web services, microservices, and interface technologies. The developed platform can be used via a standard web browser or an Application Programming Interface (API). In this study, Docker packages were used to improve system performance instead of using different virtual machines on a single server. In the experiments performed, it was shown that the Turkish ASR system had a word error rate of 24.70%. In web service performance tests, it was shown that the platform responded in an average of 9.6 seconds for a 59-second speech recording. The developed user interface was tested in both mobile and desktop web browsers and was shown to function properly. Applications and other services were given access to the platform without the need to use an interface via API support provided by the platform. As a result, a web service-based Turkish ASR platform working seamlessly on the ever-increasing number of mobile devices, the Internet of Things ecosystem, or other access devices was developed.
Main problem of energy is limited energy resources, because the energy consumption rapidly increa... more Main problem of energy is limited energy resources, because the energy consumption rapidly increases. Furthermore, efficiently use of energy is a way to reduce the problem. In this study, a smart lamp socket unit was developed to report the consumption values of the energy and to measure energy consumption in the lighting area where the consumption of electric energy is generally high. The smart lamp socket is able to communicate via wireless network connection and internet protocol. A machine to machine (M2M) platform has been developed for easily accessible user interaction, energy consumption tracking, management and reporting thanks to the smart lamp socket. Using the M2M platform, many smart lamp socket unit can be communicated at the same time. Thus any costumer can directly manage to save energy according to reported consumption data and observed the energy consumption in the lighting fields.
Author identification takes an important place in Natural Language Processing (NLP). Each written... more Author identification takes an important place in Natural Language Processing (NLP). Each written document carries the trail of its author. In this study, we aim to realize the author identification via the traces belonging to author be retrieved from the text. A raw dataset was created with 25 columnists and randomly selected 2024 texts from different newspapers in the Turkish language. A dataset with character and lexical features with natural language processing methods were prepared over the raw dataset. The feature selection process was realized with the combination of the Chicken swarm optimization and the ensemble learning algorithms on the prepared dataset. The results were evaluated before and after the feature selection method was applied. The highest success rate with 93.99% was achieved when Adaboost with J48 algorithm was applied after the feature selection process carried out.
Finding a vacant place to park cars in the rush hour is time-consuming and even may be frustratin... more Finding a vacant place to park cars in the rush hour is time-consuming and even may be frustrating for the drivers. In some studies, vehicles are equipped with communicative tools called On Board Units (OBUs), which come along with roadside devices known as Roadside Units (RSUs), which allow the drivers to communicate each other and trace a vacant parking place easily. Previous systems work with connection of sensors all over the road and parking space and may result in spending much time to find a parking place, occupancy of the empty parking place until the car gets to the desired location, requirement for an additional hardware, wired network communication, and security issues. In this paper, we propose an exponential smoothing and multi-objective decision-making by using cloud-based methods to find the best parking place, taking into consideration of the park-cost for the driver. The proposed system uses the cellular base stations to eliminate the cost of the RSUs and the sensors. The results of our simulations via NS-2 network simulator confirm the efficiency of the proposed model.
Transition from classical networks to smart grid (SGs) is important both in terms of performance ... more Transition from classical networks to smart grid (SGs) is important both in terms of performance and failure intervention. When this transition is made, the existing network structure must be transformed into a structure suitable for the SGs. The transformation in the physical structure is also required to make the collected data a single centered. One of the most important factors that will improve performance in reading and writing data in systems used in SGs or power plants is the database system. However, when considering the size of the existing grids, the multiplicity of collected data and the speed at which data is read and written are major problems. In order to solve this problem, wind energy based smart grid scenario has been established. In the created system, relational and non-relational databases are discussed. In order to test the databases, different scenarios have been applied in the developed system and the most suitable database structure has been tried to be determined.
Turkish Journal of Electrical Engineering and Computer Sciences, Mar 30, 2021
Sign languages are nonverbal, visual languages that hearing-or speech-impaired people use for com... more Sign languages are nonverbal, visual languages that hearing-or speech-impaired people use for communication. Aside from hands, other communication channels such as body posture and facial expressions are also valuable in sign languages. As a result of the fact that the gestures in sign languages vary across countries, the significance of communication channels in each sign language also differs. In this study, representing the communication channels used in Turkish sign language, a total of 8 different data streams-4 RGB, 3 pose, 1 optical flow-were analyzed. Inception 3D was used for RGB and optical flow; and LSTM-RNN was used for pose data streams. Experiments were conducted by merging the data streams in different combinations, and then a sign language recognition system that merged the most suitable streams with the help of a multistream late fusion mechanism was proposed. Considering each data stream individually, the accuracies of the RGB streams were between 28% and 79%; pose stream accuracies were between 9% and 50%; and optical flow data accuracy was 78.5%. When these data streams were used in combination, the sign language recognition performance was higher in comparison to any of the data streams alone. The proposed sign language recognition system uses a multistream data fusion mechanism and gives an accuracy of 89.3% on BosphorusSign General dataset. The multistream data fusion mechanisms have a great potential for improving sign language recognition results.
2019 3rd International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT), Oct 1, 2019
The traditional network used today is unable to meet the increasing needs of technology in terms ... more The traditional network used today is unable to meet the increasing needs of technology in terms of management, scaling, and performance criteria. Major developments in information and communication technologies show that the traditional network structure is quite lacking in meeting the current requirements. In order to solve these problems, Software Defined Network (SDN) is capable of responding as it, is flexible, easier to manage and offers a new structure. Software Defined Networks have many advantages over traditional network structure. However, it also brings along many security threats due to its new architecture. For example, the DoS attack, which overloads the controller's processing and communication capacity in the SDN structure, is a significant threat. Mobile Ad Hoc Network (MANET), which is one of the wireless network technologies, is different from SDN technology. MANET is exposed to various attacks such as DoS due to its security vulnerabilities. The aim of the study is to reveal the security problems in SDN structure presented with a new understanding. This is based on the currently used network structures such as MANET. The study consists of two parts. First, DoS attacks against the SDN controller were performed. Different SDN controllers were used for more accurate results. Second, MANET was established and DoS attacks against this network were performed. Different MANET routing protocols were used for more accurate results. According to the scenario, attacks were performed and the performance values of the networks were tested. The reason for using two different networks in this study is to compare the performance values of these networks at the time of attack. According to the test results, both networks were adversely affected by the attacks. It was observed that network performance decreased in MANET structure but there was no network interruption. The SDN controller becomes dysfunctional and collapses as a result of the attack. While the innovations offered by the SDN structure are expected to provide solutions to many problems in traditional networks, there are still many vulnerabilities for network security.
Vision-based human activity recognition in smart homes has become a significant issue in terms of... more Vision-based human activity recognition in smart homes has become a significant issue in terms of developing the next generation technologies which can improve healthcare and security of smart homes. Recently, deep learning models that aim to automatic extraction of low-level to high-level features of input data instead of using complicated conventional feature extraction methods have achieved significant improvements in the classification of a large amount of data especially vision-based datasets. Therefore, in this study in order to recognize human action of a smart home video dataset (DMLSmartActions) Convolutional Neural Networks (CNNs) architecture as a deep learning model has been proposed. Moreover, the performance of the proposed method has been compared with the previous methods which have used traditional machine learning methods on the same dataset. Experimental results demonstrated that the proposed deep learning model has achieved 82.41% accuracy rate in the classification of human activity.
Heart disease is one of the most known and deadly diseases in the world, and many people lose the... more Heart disease is one of the most known and deadly diseases in the world, and many people lose their lives from this disease every year. Early detection of this disease is vital to save people’s lives. Machine Learning (ML), an artificial intelligence technology, is one of the most convenient, fastest, and low-cost ways to detect disease. In this study, we aim to obtain an ML model that can predict heart disease with the highest possible performance using the Cleveland heart disease dataset. The features in the dataset used to train the model and the selection of the ML algorithm have a significant impact on the performance of the model. To avoid overfitting (due to the curse of dimensionality) due to the large number of features in the Cleveland dataset, the dataset was reduced to a lower dimensional subspace using the Jellyfish optimization algorithm. The Jellyfish algorithm has a high convergence speed and is flexible to find the best features. The models obtained by training the ...
Supervisory Control and Data Acquisition (SCADA) systems play a crucial role in overseeing and co... more Supervisory Control and Data Acquisition (SCADA) systems play a crucial role in overseeing and controlling renewable energy sources like solar, wind, hydro, and geothermal resources. Nevertheless, with the expansion of conventional SCADA network infrastructures, there arise significant challenges in managing and scaling due to increased size, complexity, and device diversity. Using Software Defined Networking (SDN) technology in traditional SCADA network infrastructure offers management, scaling and flexibility benefits. However, as the integration of SDN-based SCADA systems with modern technologies such as the Internet of Things, cloud computing, and big data analytics increases, cybersecurity becomes a major concern for these systems. Therefore, cyber-physical energy systems (CPES) should be considered together with all energy systems. One of the most dangerous types of cyber-attacks against SDN-based SCADA systems is Distributed Denial of Service (DDoS) attacks. DDoS attacks disrupt the management of energy resources, causing service interruptions and increasing operational costs. Therefore, the first step to protect against DDoS attacks in SDN-based SCADA systems is to develop an effective intrusion detection system. This paper proposes a Decision Tree-based Ensemble Learning technique to detect DDoS attacks in SDN-based SCADA systems by accurately distinguishing between normal and DDoS attack traffic. For training and testing the ensemble learning models, normal and DDoS attack traffic data are obtained over a specific simulated experimental network topology. Techniques based on feature selection and hyperparameter tuning are used to optimize the performance of the decision tree ensemble models. Experimental results show that feature selection, combination of different decision tree ensemble models, and hyperparameter tuning can lead to a more accurate machine learning model with better performance detecting DDoS attacks against SDN-based SCADA systems.
Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım ve Teknoloji, 2018
Günümüzde giderek artan bir oranda makinelerin birbirleri ile iletişimde olduğu görülmektedir. He... more Günümüzde giderek artan bir oranda makinelerin birbirleri ile iletişimde olduğu görülmektedir. Herhangi bir sektör için geliştirilen makineler arası iletişim (M2M: Machine to Machine) uygulamasının kullanıcı ile etkileşimi için bir M2M platformuna ihtiyaç duyulmaktadır. M2M platformunun amacı veri toplama, veri izleme, veri yönetimi ve raporlama-analiz gibi işlevleri sağlamaktır. M2M platformları çoğunlukla web tabanlı olarak geliştirilir ve kullanıcının her ortamdan platforma erişim sağlaması hedeflenir. Bu hedef doğrultusunda bazı M2M platformları hem masaüstü hem de mobil uygulama sürümleri oluşturularak kullanıcıya sunulmaktadır. Fakat bu durum mobil işletim sistemi ve mobil ekran boyutuna göre farklı arayüz tasarımları yapmayı gerektirmektedir. Bu durum arayüz geliştiricilerinin işini oldukça zorlaştırmaktadır. Bu çalışmada her mobil işletim sistemi ve ekran boyutuna göre arayüz geliştirmek yerine duyarlı tasarıma sahip web tabanlı bir M2M platformu geliştirilmiştir. Bu sayede ...
Word error rate is high when the acoustic model used in Turkish Automatic Speech Recognition (ASR... more Word error rate is high when the acoustic model used in Turkish Automatic Speech Recognition (ASR) systems is not supported by a robust language model. The use of a welldesigned language model in combination with an acoustic model in the ASR reduces the word error rate. In the training of the language model required for ASR, plain text data is used. The accuracy of the text data used is very important for the training of ASR models. In this study, the detection and correction of spelling errors in the text data used in the education of Turkish ASR system were realized by using a method based on natural language processing. Firstly, misspelled words were found in the text data. Lack of characters, surplus characters,
Türkiye Bilişim Vakfı Bilgisayar Bilimleri ve Mühendisliği Dergisi, Jul 6, 2017
Teknolojinin ve internet alt yapısının hızlı gelişimi, gömülü cihazları kullanarak insanların ile... more Teknolojinin ve internet alt yapısının hızlı gelişimi, gömülü cihazları kullanarak insanların iletişim kurma şeklini değiştirmiştir. Mevcut durumda fiziksel nesnelerin büyük bir kısmının iletişim yeteneği bulunmamaktadır. Nesnelerin interneti (IoT: Internet of Things) yaklaşımı ve Makinalar Arası İletişim (M2M: Machine to Machine) teknolojileri sayesinde nesnelere iletişim yeteneği kazandırılabilmektedir. Bu sayede nesnelerin uzaktan izlenmesi ve kontrolü gerçekleştirilmektedir. M2M/IoT uygulamalarının yönetilebilmesi için M2M/IoT platformları kullanılmaktadır. Bu çalışmada, öncelikle M2M/IoT teknolojisi açıklanmış M2M/IoT platformları üzerinde prototip uygulama geliştirilmiştir. Kullanılan platformlar mevcutta ticari olarak hizmet vermektedir. Bu platformların genel yapısı ve mimarisi, hangi alanlarda kullanıldığı, platform üzerinde uygulamaların nasıl geliştirildiği incelenmiştir. Uygulamaya bağlı olarak M2M/IoT platformlarının farklı işlevsel özelliklere sahip oldukları görülmüştür.
Machine to machine (M2M) is a smart ecosystem that enables devices to communicate by exchanging d... more Machine to machine (M2M) is a smart ecosystem that enables devices to communicate by exchanging data through specific protocols autonomously with limited human interaction (or no human interaction) between each other. Use cases, business models, varieties of devices and protocols of M2M systems getting proliferating. Being able to accommodate this large spectrum of possibilities requires a standardized M2M platform design based on abstraction, modularity, and scalability. In this study, a web-based scalable M2M platform was developed which is compatible with European Telecommunications Standards Institute (ETSI) and oneM2M standards. The web-based M2M service platform has with a sensitive design adopting Web 3.0 standards and Service Oriented Architecture (SOA) approaches. The access to M2M services and data was provided using Representational State Transfer, RestFul Web Services (REST) through a service layer. The data acquired from M2M devices were stored in a non-relational database. The security of M2M platform was achieved through token based ID check. Fast, scalable, cost-efficient, secure, standardized, general purpose an M2M service platform was created.
Uluslararası bilimsel araştırmalar dergisi, Aug 1, 2019
Respiratory sounds are perceived as the most important diagnostic value by many physicians in dia... more Respiratory sounds are perceived as the most important diagnostic value by many physicians in diagnosing patients suffering from lung disease and monitoring their clinical course. It is no doubt that the results of visual-acoustic analysis of lung sounds in the computer will contribute to the diagnosis of lung diseases. Acoustic analysis of respiration can quantify changes in lung sounds, make permanent recordings of lung sounds and produce graphic representations. Thus, monitoring and diagnosis of patients suffering from common lung diseases can be facilitated. In this study, time amplitude waveform, time expanded waveform, RMS values, frequency changes cohesive to zero crossing velocity were observed in the time domain. At the same time, thanks to the algorithms developed in the time domain, the wheeze and crackle sounds from the lung sounds were determined automatically. According to pre-defined amplitude and time criteria when an unusual event that represents wheeze and crackle sounds in various parts of the lung sounds, the detection of unusual event was performed. As a result, time-domain analysis of lung sounds has been shown to provide more information on the diagnosis of lung diseases in a non-invasive method.
The objective of this study is to investigate the effects of machining parameters on the acoustic... more The objective of this study is to investigate the effects of machining parameters on the acoustic frequency developed during the machining of cold work tool steels. The machining tests were carried out through a milling method with different machining parameters. Analogue acoustic signals obtained via the microphone were converted and digitized. During idle operation and cutting operation, the obtained spindle sound recordings were subjected to Fast Fourier Transform (FFT) and converted from the time domain to the frequency domain, and the statistical effects of cutting parameters were investigated by use of analysis of variance. Cutting speed was found to be the only influential factor for the acoustic frequency in idle time. In the starting of machining, the feed rate, cutting speed, and depth of cut were seen to affect the acoustic frequency, while the cutting speed, insert material, insert radius, depth of cut, and feed rate were seen to be effective in the later stages of machining.
In this study, developed a platform that performs content analysis on the YouTube videos. With th... more In this study, developed a platform that performs content analysis on the YouTube videos. With the developed platform, it is aimed to identify the videos supporting the terrorist organizations on YouTube and to remove the detected videos. In the study, an image analysis model was constructed based on the Convolutional Neural Network (CNN) and the AlexNet model. Were made separately analysis performed on each of the frame forming the video image. For the analysis of the images contained in the video, training data set is prepared from different YouTube videos. The analysis platform was developed with a RESTful web service based architecture using Spring Framework. A web interface for users to interact with the platform has been developed. Each user can receive analysis results of different videos via email notification. For platform testing, videos with and without terrorist propaganda were used. In the tests conducted after CNN network training, 70% accuracy rate was achieved. By increasing the number of content in the training data set of video analysis process it is thought to increase the success rate.
Bu calismada lazer ve cizgisel CCD(Charge-Coupled Device) algilayici tabanli bir optik cap olcum ... more Bu calismada lazer ve cizgisel CCD(Charge-Coupled Device) algilayici tabanli bir optik cap olcum sistemi gelistirilmistir. Bu sekilde bir olcum sistemi ile optik temassiz cap olcumleri gerceklestirilmistir. Bir lazer diyot ve kollimator lens kullanarak paralel bir isik demeti olusturulmustur. Paralel isik demeti karsisina cizgisel CCD algilayici yerlestirilmis ve lazer isik demeti ile aydinlatilmistir. Paralel isik demeti icerisine bir nesne konumlandirilip cizgisel CCD algilayici uzerine nesne golgesi dusurulmustur. Bu sekilde cizgisel CCD algilayicida golge altinda kalan kisimlardaki pikseller aydinlanmadigindan, nesnenin capi olusan golgeden belirlenebilmektedir. Calismada silindirik kesit alanli nesnelerin capinin temassiz olarak olculmesi %0,24 - %1,34 hata payi ile gerceklestirilmistir.
Akciğer hastalıklarının tanısında yeni enstrümanlar ve akciğer ses analizleri klinik uygulamalard... more Akciğer hastalıklarının tanısında yeni enstrümanlar ve akciğer ses analizleri klinik uygulamalarda hala yaygın olarak kullanılmamaktadır. Bu çalışmada solunum problemine sahip 48 hastanın oskültasyonu deneyimli bir doktor tarafından elektronik stetoskop (Littmann Electronic Stethoscope Model 4000) kullanılarak yapıldıktan sonra akciğer sesleri stetoskoba kaydedilmiş ve bu sesler doktorun hastaya ait yorumları ile beraber hastanın sürekli takibi amacıyla bilgisayara bir hasta veritabanı olarak kaydedilmiştir. Ayrıca stetoskopta bulunan sesler bilgisayara uygun ses formatında da kaydedilerek daha sonra akustik analizi için olanak sağlanmıştır. Böylece hem gerçek zamanlı dinleme hem de bilgisayar akustik analiz metotlarının kullanılması ile akciğer seslerindeki değişimler, doktorlara normal ve anormal solunum sesi bilgisini sağlamış ve daha kolay teşhis imkanı oluşturmuştur.
In response to the similar challenges in building large-scale distributed applications and platfo... more In response to the similar challenges in building large-scale distributed applications and platforms on the Web, microservice architecture has emerged and gained a lot of popularity in recent years. Therefore, both for the use of microservices and for the provided of the necessary interface for Automatic Speech Recognition (ASR), a web-based platform has been developed. Within firstly the scope of the study, a Turkish ASR system was developed. A web service structure was created to facilitate access to the ASR system. The access of methods and data in the web service structure was provided through Representational State Transfer (REST) web services and service layer. An interface was developed to enable interaction with the web service. The platform was developed using a combination of different technologies such as ASR, web services, microservices, and interface technologies. The developed platform can be used via a standard web browser or an Application Programming Interface (API). In this study, Docker packages were used to improve system performance instead of using different virtual machines on a single server. In the experiments performed, it was shown that the Turkish ASR system had a word error rate of 24.70%. In web service performance tests, it was shown that the platform responded in an average of 9.6 seconds for a 59-second speech recording. The developed user interface was tested in both mobile and desktop web browsers and was shown to function properly. Applications and other services were given access to the platform without the need to use an interface via API support provided by the platform. As a result, a web service-based Turkish ASR platform working seamlessly on the ever-increasing number of mobile devices, the Internet of Things ecosystem, or other access devices was developed.
Main problem of energy is limited energy resources, because the energy consumption rapidly increa... more Main problem of energy is limited energy resources, because the energy consumption rapidly increases. Furthermore, efficiently use of energy is a way to reduce the problem. In this study, a smart lamp socket unit was developed to report the consumption values of the energy and to measure energy consumption in the lighting area where the consumption of electric energy is generally high. The smart lamp socket is able to communicate via wireless network connection and internet protocol. A machine to machine (M2M) platform has been developed for easily accessible user interaction, energy consumption tracking, management and reporting thanks to the smart lamp socket. Using the M2M platform, many smart lamp socket unit can be communicated at the same time. Thus any costumer can directly manage to save energy according to reported consumption data and observed the energy consumption in the lighting fields.
Author identification takes an important place in Natural Language Processing (NLP). Each written... more Author identification takes an important place in Natural Language Processing (NLP). Each written document carries the trail of its author. In this study, we aim to realize the author identification via the traces belonging to author be retrieved from the text. A raw dataset was created with 25 columnists and randomly selected 2024 texts from different newspapers in the Turkish language. A dataset with character and lexical features with natural language processing methods were prepared over the raw dataset. The feature selection process was realized with the combination of the Chicken swarm optimization and the ensemble learning algorithms on the prepared dataset. The results were evaluated before and after the feature selection method was applied. The highest success rate with 93.99% was achieved when Adaboost with J48 algorithm was applied after the feature selection process carried out.
Finding a vacant place to park cars in the rush hour is time-consuming and even may be frustratin... more Finding a vacant place to park cars in the rush hour is time-consuming and even may be frustrating for the drivers. In some studies, vehicles are equipped with communicative tools called On Board Units (OBUs), which come along with roadside devices known as Roadside Units (RSUs), which allow the drivers to communicate each other and trace a vacant parking place easily. Previous systems work with connection of sensors all over the road and parking space and may result in spending much time to find a parking place, occupancy of the empty parking place until the car gets to the desired location, requirement for an additional hardware, wired network communication, and security issues. In this paper, we propose an exponential smoothing and multi-objective decision-making by using cloud-based methods to find the best parking place, taking into consideration of the park-cost for the driver. The proposed system uses the cellular base stations to eliminate the cost of the RSUs and the sensors. The results of our simulations via NS-2 network simulator confirm the efficiency of the proposed model.
Transition from classical networks to smart grid (SGs) is important both in terms of performance ... more Transition from classical networks to smart grid (SGs) is important both in terms of performance and failure intervention. When this transition is made, the existing network structure must be transformed into a structure suitable for the SGs. The transformation in the physical structure is also required to make the collected data a single centered. One of the most important factors that will improve performance in reading and writing data in systems used in SGs or power plants is the database system. However, when considering the size of the existing grids, the multiplicity of collected data and the speed at which data is read and written are major problems. In order to solve this problem, wind energy based smart grid scenario has been established. In the created system, relational and non-relational databases are discussed. In order to test the databases, different scenarios have been applied in the developed system and the most suitable database structure has been tried to be determined.
Turkish Journal of Electrical Engineering and Computer Sciences, Mar 30, 2021
Sign languages are nonverbal, visual languages that hearing-or speech-impaired people use for com... more Sign languages are nonverbal, visual languages that hearing-or speech-impaired people use for communication. Aside from hands, other communication channels such as body posture and facial expressions are also valuable in sign languages. As a result of the fact that the gestures in sign languages vary across countries, the significance of communication channels in each sign language also differs. In this study, representing the communication channels used in Turkish sign language, a total of 8 different data streams-4 RGB, 3 pose, 1 optical flow-were analyzed. Inception 3D was used for RGB and optical flow; and LSTM-RNN was used for pose data streams. Experiments were conducted by merging the data streams in different combinations, and then a sign language recognition system that merged the most suitable streams with the help of a multistream late fusion mechanism was proposed. Considering each data stream individually, the accuracies of the RGB streams were between 28% and 79%; pose stream accuracies were between 9% and 50%; and optical flow data accuracy was 78.5%. When these data streams were used in combination, the sign language recognition performance was higher in comparison to any of the data streams alone. The proposed sign language recognition system uses a multistream data fusion mechanism and gives an accuracy of 89.3% on BosphorusSign General dataset. The multistream data fusion mechanisms have a great potential for improving sign language recognition results.
2019 3rd International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT), Oct 1, 2019
The traditional network used today is unable to meet the increasing needs of technology in terms ... more The traditional network used today is unable to meet the increasing needs of technology in terms of management, scaling, and performance criteria. Major developments in information and communication technologies show that the traditional network structure is quite lacking in meeting the current requirements. In order to solve these problems, Software Defined Network (SDN) is capable of responding as it, is flexible, easier to manage and offers a new structure. Software Defined Networks have many advantages over traditional network structure. However, it also brings along many security threats due to its new architecture. For example, the DoS attack, which overloads the controller's processing and communication capacity in the SDN structure, is a significant threat. Mobile Ad Hoc Network (MANET), which is one of the wireless network technologies, is different from SDN technology. MANET is exposed to various attacks such as DoS due to its security vulnerabilities. The aim of the study is to reveal the security problems in SDN structure presented with a new understanding. This is based on the currently used network structures such as MANET. The study consists of two parts. First, DoS attacks against the SDN controller were performed. Different SDN controllers were used for more accurate results. Second, MANET was established and DoS attacks against this network were performed. Different MANET routing protocols were used for more accurate results. According to the scenario, attacks were performed and the performance values of the networks were tested. The reason for using two different networks in this study is to compare the performance values of these networks at the time of attack. According to the test results, both networks were adversely affected by the attacks. It was observed that network performance decreased in MANET structure but there was no network interruption. The SDN controller becomes dysfunctional and collapses as a result of the attack. While the innovations offered by the SDN structure are expected to provide solutions to many problems in traditional networks, there are still many vulnerabilities for network security.
Vision-based human activity recognition in smart homes has become a significant issue in terms of... more Vision-based human activity recognition in smart homes has become a significant issue in terms of developing the next generation technologies which can improve healthcare and security of smart homes. Recently, deep learning models that aim to automatic extraction of low-level to high-level features of input data instead of using complicated conventional feature extraction methods have achieved significant improvements in the classification of a large amount of data especially vision-based datasets. Therefore, in this study in order to recognize human action of a smart home video dataset (DMLSmartActions) Convolutional Neural Networks (CNNs) architecture as a deep learning model has been proposed. Moreover, the performance of the proposed method has been compared with the previous methods which have used traditional machine learning methods on the same dataset. Experimental results demonstrated that the proposed deep learning model has achieved 82.41% accuracy rate in the classification of human activity.
Uploads
Papers by Huseyin Polat