Bilgisayar Mühendisliği Bölümü / Department of Computer Engineering
Permanent URI for this collectionhttps://hdl.handle.net/11413/6817
Browse
Browsing Bilgisayar Mühendisliği Bölümü / Department of Computer Engineering by Rights "http://creativecommons.org/licenses/by-nc-nd/3.0/us/"
Now showing 1 - 20 of 42
- Results Per Page
- Sort Options
Publication 2D UAV path planning with radar threatening areas using simulated annealing algorithm for event detection(2018) Basbous, Bilal;Path Planning for Unmanned Aerial Vehicles (UAVs) can be used for many purposes. However, the problem becomes more and more complex when dealing with a large number of points to visit for detecting and catching different type of events and simple threat avoidance such as Radar Areas. In the literature different type of algorithms (especially evolutionary algorithms) are preferred. In this project, Simulated Annealing (SA) Algorithm is used for solving the path planning problem. Firstly, problem is converted to a part of Travelling Salesman Problem (TSP), and then the solutions are optimized with the 2-Opt approach and other simple algorithms. The code is implemented in MATLAB by using its visualization. Circular avoidance approach is developed and applied with the Simulated Annealing in order to escape from circular radar threats. Tests have been made to observe the results of SA algorithm and radar threats avoidance approaches, where the results show that after a period of time, SA algorithm gives acceptable solutions with the capacities of escaping from radar area threats. Where SA algorithm gives better solutions in less period of time when there are no radar threats. Experimental results depicted that the proposed model can result in an acceptable solution for UAVs in sufficient execution time. This model can be used as an alternative solution to the similar evolutionary algorithms.Publication A Novel Input Set for LSTM based Transport Mode Detection(2019-03) Güvensan, M.Amaç; AŞCI, GÜVEN; 285689The capability of mobile phones are increasing with the development of hardware and software technology. Especially sensors on smartphones enable to collect environmental and personal information. Thus, smartphones become the key components of ambient intelligence. Human activity recognition and transport mode detection (TMD) are the main research areas for tracking the daily activities of a person. This study aims to introduce a novel input set for daily activities mainly for transportation modes in order to increase the detection rate. In this study, the frame-based novel input set consisting of time-domain and frequency-domain features are fed to LSTM network. Thus, the classification ratio on HTC public dataset is climbed up to 97% which is 2% more than the state-of-the-art method in the literature.Publication A pipeline for adaptive filtering and transformation of noisy left-arm ECG to its surrogate chest signal(MDPI AG, 2020-05) Tanneeru, Akhilesh; Lee, Bongmook; Misra, Veena; Mohaddes, F.; Zhou, Y.; Lobaton, E.; AKBULUT, FATMA PATLARThe performance of a low-power single-lead armband in generating electrocardiogram (ECG) signals from the chest and left arm was validated against a BIOPAC MP160 benchtop system in real-time. The filtering performance of three adaptive filtering algorithms, namely least mean squares (LMS), recursive least squares (RLS), and extended kernel RLS (EKRLS) in removing white (W), power line interference (PLI), electrode movement (EM), muscle artifact (MA), and baseline wandering (BLW) noises from the chest and left-arm ECG was evaluated with respect to the mean squared error (MSE). Filter parameters of the used algorithms were adjusted to ensure optimal filtering performance. LMS was found to be the most effective adaptive filtering algorithm in removing all noises with minimum MSE. However, for removing PLI with a maximal signal-to-noise ratio (SNR), RLS showed lower MSE values than LMS when the step size was set to 1 × 10−5. We proposed a transformation framework to convert the denoised left-arm and chest ECG signals to their low-MSE and high-SNR surrogate chest signals. With wide applications in wearable technologies, the proposed pipeline was found to be capable of establishing a baseline for comparing left-arm signals with original chest signals, getting one step closer to making use of the left-arm ECG in clinical cardiac evaluations.Publication A tree learning approach to web document sectional hierarchy extraction(2010) Pembe, F.Canan; Göngör, TungaThere is an increasing availability of documents in electronic form due to the widespread use of the Internet. Hypertext Markup Language (HTML) which is mostly concerned with the presentation of documents is still the most commonly used format on the Web, despite the appearance of semantically richer markup languages such as XML. Effective processing of Web documents has several uses such as the display of content on small-screen devices and summarization. In this paper, we investigate the problem of identifying the sectional hierarchy of a given HTML document together with the headings in the document. We propose and evaluate a learning approach suitable to tree representation based on Support Vector Machines.Publication A Wearable Device for Virtual Cyber Therapy of Phantom Limb Pain(2018-09) Tarakçı, Ela; Aydın, Muhammed; Zaim, Abdul Halim; AKBULUT, AKHAN; AŞCI, GÜVEN; 285689; 116056; 101760; 176402; 8693Phantom limb pain (PLP) is the condition most often occurs in people who have had a limb amputated and it is may affect their life severely. When the brain sends movement signals to the phantom limb, it returns and causes a pain. Many medical approaches aim to treat the PLP, however the mirror therapy still considered as the base therapy method. The aim of this research is to develop a wearable device that measures the EMG signals from PLP patients to classify movements on the amputated limb. These signals can be used in virtual reality and augmented reality environments to realize the movements in order to reduce pain. A data set was generated with measurements taken from 8 different subjects and the classification accuracy achieved as 90% with Neural Networks method that can be used in cyber therapies.This type of therapy provides strong visuals which make the patient feel he/she really have the limb. The patient will have great therapy session time with comparison to the other classical therapy methods that can be used in home environments.Publication Analysis of Facial Emotion Expression in Eating Occasions Using Deep Learning(Springer, 2023) ELİF, YILDIRIM; AKBULUT, FATMA PATLAR; Çatal, ÇağatayEating is experienced as an emotional social activity in any culture. There are factors that influence the emotions felt during food consumption. The emotion felt while eating has a significant impact on our lives and affects different health conditions such as obesity. In addition, investigating the emotion during food consumption is considered a multidisciplinary problem ranging from neuroscience to anatomy. In this study, we focus on evaluating the emotional experience of different participants during eating activities and aim to analyze them automatically using deep learning models. We propose a facial expression-based prediction model to eliminate user bias in questionnaire-based assessment systems and to minimize false entries to the system. We measured the neural, behavioral, and physical manifestations of emotions with a mobile app and recognize emotional experiences from facial expressions. In this research, we used three different situations to test whether there could be any factor other than the food that could affect a person’s mood. We asked users to watch videos, listen to music or do nothing while eating. This way we found out that not only food but also external factors play a role in emotional change. We employed three Convolutional Neural Network (CNN) architectures, fine-tuned VGG16, and Deepface to recognize emotional responses during eating. The experimental results demonstrated that the fine-tuned VGG16 provides remarkable results with an overall accuracy of 77.68% for recognizing the four emotions. This system is an alternative to today’s survey-based restaurant and food evaluation systems.Publication Automatic HTML code generation from mock-up images using machine learning techniques(2019) Asiroğlu, Batuhan; Yıldız, Eyyüp; Nalçakan, Yağız; Sezen, Alper; Dağtekin, Mustafa; Ensari, Tolga; METE, BÜŞRA RÜMEYSAThe design cycle for a web site starts with creating mock-ups for individual web pages either by hand or using graphic design and specialized mock-up creation tools. The mock-up is then converted into structured HTML or similar markup code by software engineers. This process is usually repeated many more times until the desired template is created. In this study, our aim is to automate the code generation process from hand-drawn mock-ups. Hand drawn mock-ups are processed using computer vision techniques and subsequently some deep learning methods are used to implement the proposed system. Our system achieves 96% method accuracy and 73% validation accuracy.Publication Autonomous vehicle control for Lane and vehicle tracking by using deep learning via vision(2018) Olgun, Masum Celil; Baytar, Zakir; Akpolat, Kadir Metin; ŞAHİNGÖZ, ÖZGÜR KORAYCamera-based lane detection and vehicle tracking algorithms are one of the keystones for many autonomous systems. The navigational process of those systems is mainly focused on the output of detection algorithms. However, detection algorithms for lane detection need more pre-processing time and computational effort. They are also affected by environmental conditions and must regularly be improved. In this paper machine learning techniques and computer vision algorithms are utilized for the tasks of the lane and vehicle tracking of an autonomous vehicle control scenario. With the nature of used learning algorithm, the proposed system can handle complex image problems. The vehicle, on which we implement our algorithms, can manage to carry out the following tasks autonomously; tracking the lanes, following another vehicle, and stopping in necessary conditions. For that, one of the primary purposes is image-based lane tracking methodology by using learning algorithms. Data augmentation is applied to create diversity for the dataset. Application in this methodology has been discussed. For lane tracking Convolutional Neural Network architecture which is based on NVIDIA's PilotNet is preferred. For detecting objects and vehicles, the system is trained on the faster region-based convolutional neural network (Faster R-CNN) to identify traffic light and stop sign are by Haar Cascade Classifier. All these learning models are trained on NVIDIA GTX 1070 Graphics Processing Unit (GPU) to reduce training time. Experimental results showed that the proposed system gives a favorable result to autonomously control vehicles for lane and vehicle tracking purposes by vision.Publication Bimodal affect recognition based on autoregressive hidden Markov models from physiological signals(2020-10) Perros, H.G; Shahzad, M.; AKBULUT, FATMA PATLARBackground and objective: Affect provides contextual information about the emotional state of a person as he/she communicates in both verbal and/or non-verbal forms. While human's are great at determining the emotional state of people while they communicate in person, it is challenging and still largely an unsolved problem to computationally determine the emotional state of a person. Methods: Emotional states of a person manifest in the physiological biosignals such as electrocardiogram (ECG) and electrodermal activity (EDA) because these signals are impacted by the peripheral nervous system of the body, and the peripheral nervous system is strongly coupled with the mental state of the person. In this paper, we present a method to accurately recognize six emotions using ECG and EDA signals and applying autoregressive hidden Markov models (AR-HMMs) and heart rate variability analysis on these signals. The six emotions include happiness, sadness, surprise, fear, anger, and disgust. Results: We evaluated our method on a comprehensive new dataset collected from 30 participants. Our results show that our proposed method achieves an average accuracy of 88.6% in distinguishing across the 6 emotions. Conclusions: The key technical depth of the paper is in the use of the AR-HMMs to model the EDA signal and the use of LDA to enable accurate emotion recognition without requiring a large number of training samples. Unlike other studies, we have taken a hierarchical approach to classify emotions, where we first categorize the emotion as either positive or negative and then identify the exact emotion.Publication Boosting the Visibility of Services in Microservice Architecture(Springer, 2023) TOKMAK, AHMET VEDAT; AKBULUT, AKHAN; Çatal, ÇağatayMonolithic software architectures are no longer sufficient for the highly complex software-intensive systems, which modern society depends on. Service Oriented Architecture (SOA) surpassed monolithic architecture due to its reusability, platform independency, ease of maintenance, and scalability. Recent SOA implementations made use of cloud-native architectural approaches such as microservice architecture, which has resulted in a new challenge: the discovery difficulties of services. One way to dynamically discover and route traffic to service instances is to use a service discovery tool to locate the Internet Protocol (IP) address and port number of a microservice. In the event that replicated microservice instances are found to provide the same function, it is crucial to select the right microservice that provides the best overall experience for the end-user. Parameters including success rate, efficiency, delay time, and response time play a vital role in establishing a microservice's Quality of Service (QoS). These assessments can be performed by means of a live health-check service, or, alternatively, by making a prediction of the current state of affairs with the application of machine learning-based approaches. In this research, we evaluate the performance of several classification algorithms for estimating the quality of microservices using the QWS dataset containing traffic data of 2505 microservices. Our research also analyzed the boosting algorithms, namely Gradient Boost, XGBoost, LightGBM, and CatBoost to improve the overall performance. We utilized parameter optimization techniques, namely Grid Search, Random Search, Bayes Search, Halvin Grid Search, and Halvin Random Search to fine-tune the hyperparameters of our classifier models. Experimental results demonstrated that the CatBoost algorithm achieved the highest level of accuracy (90.42%) in predicting microservice quality.Publication Cases of Zika virus infection in Turkey: newly married couple returning from Cuba(2018-07) Alper, Sezen; Yıldırım, M.; Kültür, MN.; Pehlivanoğlu, F.; Menemenlioğlu, D.Zika virus is a Flavivirus in the family Flaviviridae, and transmitted to humans by Aedes species mosquitoes. Zika virus infection is asymptomatic in 80% of cases and has a mild course when symptoms occur. These symptoms include headache, myalgia, mild fever, maculopapular rash and conjunctivitis. Zika virus has been associated with serious neurological complications such as Guillain-Barre syndrome in adults and microcephaly development in neonates. It has recently become a global public health problem as a result of increasing cases. As it is known that the vector of this disease is present in our country; entry of Zika virus infection in our country has a great importance. In this report the clinical and laboratory findings of two cases of Zika virus infection imported to Turkey by a couple returning from Cuba in October 2017 were presented. Newly married couple, both 29 years old, without a known chronic disease history, went on honeymoon to Cuba between 17-29 September and they visited Havana and Varadero. They reported that they were bitten repeatedly by the mosquitoes and did not use insect repellents during this time. Four days after returning to Turkey, they had headaches, back pain and myalgia followed by rash and joint pain. They reported having their symptoms started in the same day in a few hours difference. The symptoms for both patients disappeared in 10 days. Serum samples from the patients were sent to the Public Health General Directorate National Arboviruses and Viral Zoonoses Laboratory to be tested for Dengue, Chikungunya and Zika viruses. Nucleic acid testing yielded negative results. The Arbovirus Indirect Immunofluorescence test were positive both for IgM and IgG for Zika virus. No cross reactivity with Dengue virus was detected. Chikungunya antibodies were found as negative. At two months of the diagnosis, urine and semen samples of the male patient were tested by real-time reverse transcriptase polymerase chain reaction (rRT-PCR). The result was negative for urine but positive for semen sample. This report is important to present the first cases of Zika virus infection published in Turkey. Zika virus infection should be suspected in patients with fever, headache, rash, myalgia and joint pain returning from an endemic areas. All travelers, especially pregnant women, have to take precautions for mosquitos during the trip.Publication Code Generator Framework for Smart TV Platforms(INST ENGINEERING TECHNOLOGY-IET, MICHAEL FARADAY HOUSE SIX HILLS WAY STEVENAGE, HERTFORD SG1 2AY, ENGLAND, 2019-08) TOPRAK, SEZER; AKBULUT, AKHAN; 116056; 176006In recent years, smart TVs have become more common, making them need to be included as targets for the software industry. In this study, the authors developed a code generator framework and demonstrated it in an architectural view. The proposed framework converts C# programming language based projects, in a Windows Forms or a Windows Phone Application project, into native smart TV Platform applications. The selected primary smart TV platforms assigned for application conversion were Android TV, Firefox OS, and Tizen OS. The authors enabled developers to generate native codes for all three platforms from a single code base using model to model conversion, as in the model driven architecture approach with the use of the open source Roslyn C# language compiler. The need for creating projects for every single platform to make them run on different platforms will thus be eliminated and development cycles shortened. By doing so, the time required to develop an application for each platform is reduced while keeping the generated applications' quality as high as the original application. To show the functionality, the proposed approach is applied in three case studies. The success of the code conversion is satisfactory and converted applications are functional.Publication Comparison of Cryptography Algorithms for Mobile Payment Systems(2018-10) Aydın, Muhammed; Sertbaş, Ahmet; ŞENGEL, ÖZNUR; 144004; 187909; 2285Mobile payment services are the newest and most popular technology that is developing according to our habits and needs. Consumer all over the world are using mobile phone for payment as well as communication. The main purpose of using mobile payment application is doing all transaction easily and quickly. Not only data security in electronic transactions, but also the speed of the system operations is becoming very important. There is a threshold value to finish all transaction in mobile payment systems. If the security algorithm is more complex and exceed threshold, it is not suitable to using in mobile payment systems. In this paper we compare cryptography algorithms and proposed two algorithms on Advanced Encryption Standards. The experiment results show that proposed algorithms is suitable cryptography algorithm for mobile system according to time and storage consumption factors.Publication Consensual classification of drug and nondrug compounds(2008) Pehlivanlı, Ayça C.; İbrikçi, Turgay; Ersoy, Okan K.A special consensual approach is discussed for separating a molecular group with a proven pharmacological activity from another molecular group without any activity. It is mainly a group decision to produce a consensus of multiple classification results obtained with a single classification algorithm. For this purpose, the constructed model has a preprocessing unit which consists of transformation of input patterns by random matrices and median filtering to generate independent errors for a single type of classifier and postprocessing for consensus. The neural network based consensus classifier operating with MOE descriptors was applied to a set of 641 chemical structures. The confirmed drugs were classified with an accuracy of 86.54% while nondrugs resulted in 82.67% accuracy.Publication Content-based publish/subscribe communication model between Iot devices in smart city environment(2019-04) Öztürk, Fulya; Özdemir, Ayşe MelihaIn recent years the population of the cities has been increasing the getting over half of the whole world population. These peoples are facing with some security and infrastructural needs. Some of these needs can be met with the use of some smart technologies such as Internet of Things Devices (IoT) with different types of sensors. Coordination and communication of these devices are very critical for enabling the digital solutions in a secure and comfortable city life. However, due to the restricted ability of these devices, setting up a flexible communication platform is a very challenging issue. In this paper for setting up the physical security of smart homes/buildings, a content-based publish-subscribe model is proposed by the use of wireless sensor network nodes in the critical environment. Due to the energy constraint of the devices, the broadcasting feature of the wireless communication is not used, instead of this a peer to a communication according to the content of the message is preferred. Experimental results showed that the proposed communication model can be applicable for smart city environments.Publication Deep Learning Approaches for Phantom Movement Recognition(2019-11-03) Güngör, Faray; Tarakçı, Ela; Aydın, Muhammed Ali; Zaim, Abdül Halim; AKBULUT, AKHAN; AŞCI, GÜVEN; 116056; 285689; 277179; 101760; 176402; 8693Phantom limb pain has a negative effect on the life of individuals as a frequent consequence of limb amputation. The movement ability on the lost extremity can still be maintained after the amputation or deafferentation, which is called the phantom movement. The detection of these movements makes sense for cybertherapy and prosthetic control for amputees. In this paper, we employed several deep learning approaches to recognize phantom movements of the three different amputation regions including above-elbow, below-knee and above-knee. We created a dataset that contains 25 healthy and 16 amputee participants’ surface electromyography (sEMG) readings via a wearable device with 2-channel EMG sensors. We compared the results of three different deep learning methods, respectively, Multilayer Perceptron, Convolutional Neural Network, and Recurrent Neural Network with the accuracies of two well-known shallow methods, k Nearest Neighbor and Random Forest. Our experiments indicate, Convolutional Neural Network-based model achieved an accuracy of 74.48% in recognizing phantom movements of amputees.Publication Deep Learning Based Document Modeling for Personality Detection from Turkish Texts(2019-10-24) İLGEN, BAHAR; 141812The usage of social media is increasing exponentially since it has been the easiest and fastest way to share information between people or organizations. As a result of this broad usage and activity of people on social networks, considerable amount of data is generated continuously. The availability of user generated data makes it possible to analyze personality of people. Personality is the most distinctive feature for an individual. The results of these analyses can be utilized in several ways. They provide support for human resources recruitment units to consider suitable candidates. Similar products and services can be offered to people who share the similar personality characteristics. Personality traits help in diagnosis of certain mental illnesses. It is also helpful in forensics to use personality traits on suspects to clarify the forensic case. With the rapid dissemination of online documents in many different languages, the classification of these documents has become an important requirement. Machine Learning (ML) and Natural Language Processing (NLP) methods are used to classify these digitized data. In this study, current ML techniques and methodologies have been used to classify text documents and analyze person characteristics from these datasets. As a result of classification, detailed information about the personality traits of the writer could be obtained. It was understood that the frequency-based analysis and the use of the emotional words at the word level are very important in the textual personality analysis.Publication Deep learning based forecasting in stock market with big data analytics(2019) Şişmanoğlu, Gözde; Önde, Mehmet Ali; Koçer, Furkan; ŞAHİNGÖZ, ÖZGÜR KORAYIn recent years, due to the technological improvements in computers' hardware and enhancements in the machine learning techniques, there are two increasing approaches for problem-solving as the use of "Big Data" and "Parallel Processing". Especially with the emergence of Deep Learning algorithms which can be executed parallelly on multi-core computing devices such as GPUs and CPUs, lots of real-world problems are resolved with these approaches. One of the most critical application areas in the Financial Market especially sits on Stock Markets. In this area, the aim is trying to predict the future value of a specific stock by looking at its previous financial data on the exchange process in the market. In this paper, we proposed a system that uses a Deep Learning based approach for training and constructing a knowledge base on a specific stock such as "IBM". We get time series values of the stock from the New York Stock Exchange which starts from 1968 up to 2018. Experimental results showed that this approach produces very good forecasting for specific stocks.Publication Deep learning based security management of information systems: A comparative study(2020-01) Çebi, Cem Berke; Bulut, Fatma Sena; Fırat, Hazal; ŞAHİNGÖZ, ÖZGÜR KORAY; BAYDOĞMUŞ, GÖZDE KARATAŞ; 214903In recent years, there is a growing trend of internetization which is a relatively new word for our global economy that aims to connect each market sectors (or even devices) by using the global network architecture as the Internet. Although this connectivity enables great opportunities in the marketplace, it results in many security vulnerabilities for admins of the computer networks. Firewalls and Antivirus systems are preferred as the first line of a defense mechanism; they are not sufficient to protect the systems from all type of attacks. Intrusion Detection Systems (IDSs), which can train themselves and improve their knowledge base, can be used as an extra line of the defense mechanism of the network. Due to its dynamic structure, IDSs are one of the most preferred solution models to protect the networks against attacks. Traditionally, standard machine learning methods are preferred for training the system. However, in recent years, there is a growing trend to transfer these standard machine learning-based systems to the deep learning models. Therefore, in this paper, IDSs with four different deep learning models are proposed, and their performance is compared. The experimental results showed that proposed models result in very high and acceptable accuracy rates with KDD Cup 99 Dataset.Publication Design and Implementation of a Deep Learning-Empowered m-Health Application(Springer, 2023) AKBULUT, AKHAN; DESOUKI, SARA; ABDELKHALIQ, SARA; KHANTOMANI, LAYAL; Çatal, ÇağatayMany people are unaware of the severity of melanoma disease even though such a disease can be fatal if not treated early. This research aims to facilitate the diagnosis of melanoma disease in people using a mobile health application because some people do not prefer to visit a dermatologist due to several concerns such as feeling uncomfortable by exposing their bodies. As such, a skincare application was developed so that a user can easily analyze a mole at any part of the body and get the diagnosis results quickly. In the first phase, the corresponding image is extracted and sent to a web service. Later, the web service classifies using the pre-trained model built based on a deep learning algorithm. The final phase displays the confidence rates on the mobile application. The proposed model utilizes the Convolutional Neural Network and provides 84% accuracy and 72% precision. The results demonstrate that the proposed model and the corresponding mobile application provide remarkable results for addressing the specified health problem.
- «
- 1 (current)
- 2
- 3
- »