Yazar "Unlersen, Muhammed Fahri" seçeneğine göre listele

Listeleniyor 1 - 8 / 8
  • Küçük Resim Yok
    Öğe
    ABC-ANN Based Indoor Position Estimation Using Preprocessed RSSI
    (Mdpi, 2022) Unlersen, Muhammed Fahri
    The widespread use of mobile devices has popularized the idea of indoor navigation. The Wi-Fi fingerprint method is emerging as an important alternative indoor positioning method for GPS usage difficulties. This study utilizes RSSI signals with three preprocessed states (raw, preprocessed with the path loss adapted, and exponential transformed) to train and test an artificial neural network (ANN). A systematic approach to the determination of neuron numbers in the hidden layers and activation functions of ANN is provided. The ANN is trained by the artificial bee colony algorithm. Five ML methods have been employed for estimation. The best performance has been achieved with ABC-ANN by the path loss adapted database with the MAE of 1.01 m. The estimation done using processed RSSI values has better performance than raw RSSI values. In addition, 33% less error occurs with the mentioned method compared to the data set source study.
  • Küçük Resim Yok
    Öğe
    CNN-based transfer learning-BiLSTM network: A novel approach for COVID-19 infection detection
    (Elsevier, 2021) Aslan, Muhammet Fatih; Unlersen, Muhammed Fahri; Sabanci, Kadir; Durdu, Akif
    Coronavirus disease 2019 (COVID-2019), which emerged in Wuhan, China in 2019 and has spread rapidly all over the world since the beginning of 2020, has infected millions of people and caused many deaths. For this pandemic, which is still in effect, mobilization has started all over the world, and various restrictions and precautions have been taken to prevent the spread of this disease. In addition, infected people must be identified in order to control the infection. However, due to the inadequate number of Reverse Transcription Polymerase Chain Reaction (RT-PCR) tests, Chest computed tomography (CT) becomes a popular tool to assist the diagnosis of COVID-19. In this study, two deep learning architectures have been proposed that automatically detect positive COVID-19 cases using Chest CT X-ray images. Lung segmentation (preprocessing) in CT images, which are given as input to these proposed architectures, is performed automatically with Artificial Neural Networks (ANN). Since both architectures contain AlexNet architecture, the recommended method is a transfer learning application. However, the second proposed architecture is a hybrid structure as it contains a Bidirectional Long Short-Term Memories (BiLSTM) layer, which also takes into account the temporal properties. While the COVID-19 classification accuracy of the first architecture is 98.14%, this value is 98.70% in the second hybrid architecture. The results prove that the proposed architecture shows outstanding success in infection detection and, therefore this study contributes to previous studies in terms of both deep architectural design and high classification success. (C) 2020 Elsevier B.V. All rights reserved.
  • Küçük Resim Yok
    Öğe
    CNN-SVM hybrid model for varietal classification of wheat based on bulk samples
    (Springer, 2022) Unlersen, Muhammed Fahri; Sonmez, Mesut Ersin; Aslan, Muhammet Fatih; Demir, Bedrettin; Aydin, Nevzat; Sabanci, Kadir; Ropelewska, Ewa
    Determining the variety of wheat is important to know the physical and chemical properties which may be useful in grain processing. It also affects the price of wheat in the food industry. In this study, a Convolutional Neural Network (CNN)-based model was proposed to determine wheat varieties. Images of four different piles of wheat, two of which were the bread and the remaining durum wheat, were taken and image pre-processing techniques were applied. Small-sized images were cropped from high-resolution images, followed by data augmentation. Then, deep features were extracted from the obtained images using pre-trained seven different CNN models (AlexNet, ResNet18, ResNet50, ResNet101, Inceptionv3, DenseNet201, and Inceptionresnetv2). Support Vector Machines (SVM) classifier was used to classify deep features. The classification accuracies obtained by classification with various kernel functions such as Linear, Quadratic, Cubic and Gaussian were compared. The highest wheat classification accuracy was achieved with the deep features extracted with the Densenet201 model. In the classification made with the Cubic kernel function of SVM, the accuracy value was 98.1%.
  • Küçük Resim Yok
    Öğe
    A convolutional neural network-based comparative study for pepper seed classification: Analysis of selected deep features with support vector machine
    (Wiley, 2022) Sabanci, Kadir; Aslan, Muhammet Fatih; Ropelewska, Ewa; Unlersen, Muhammed Fahri
    The seeds of high quality are very important for the cultivation of the pepper. The required cultivation practices and growing conditions may be affected by the cultivar. Also, the productivity and properties of pepper depend on the cultivar. The selection of appropriate seed cultivars may be necessary for the breeding programs. The cultivar differentiation of pepper seeds may be tested by the human eye. However, small sizes and visual similarities make it difficult to distinguish between seed cultivars. Computer vision and artificial intelligence can provide high cultivar discrimination accuracy and the procedures are objective and fast. This study aimed to classify pepper seeds belonging to different cultivars with convolutional neural network (CNN) models. The seeds were obtained from green, orange, red, and yellow pepper cultivars. A flatbed scanner was used to acquire the pepper seed images. After the image acquisition, the procedure applied was preprocessing of the images, data augmentation using different techniques and then deep learning-based classification. Two approaches have been proposed for classification. In the first approach, CNN models (ResNet18 and ResNet50) were trained for pepper seeds. In the second approach, different from the first, the features of pretrained CNN models were fused, and feature selection was applied to the fused features. Classification using all features and selected features was performed with the support vector machine (SVM) with different kernel functions (Linear, Quadratic, Cubic, Gaussian). The accuracies in the first approximation were 98.05% and 97.07% for ResNet50 and ResNet18, respectively. In the second approach, CNN-SVM-Cubic achieved up to 99.02% accuracy with the selected features. Practical applications In precision agriculture, it is very important that the seeds be of the same type for the purification and standardization of the crop culture. Performing this classification manually with human assistance will result in subjective, slow, and low standard outcomes. To overcome such problems, classification supported by artificial intelligence and machine vision systems emerges as an important tool. In this study, a highly successful classification system is presented according to the visual characteristics of pepper seeds. The proposed models can be preferred in practice for identifying pepper seeds and detecting falsification or ensuring their reliability. It will prevent mixing of different pepper seeds with different attributes for processing.
  • Küçük Resim Yok
    Öğe
    COVID-19 diagnosis using state-of-the-art CNN architecture features and Bayesian Optimization
    (Pergamon-Elsevier Science Ltd, 2022) Aslan, Muhammet Fatih; Sabanci, Kadir; Durdu, Akif; Unlersen, Muhammed Fahri
    The coronavirus outbreak 2019, called COVID-19, which originated in Wuhan, negatively affected the lives of millions of people and many people died from this infection. To prevent the spread of the disease, which is still in effect, various restriction decisions have been taken all over the world. In addition, the number of COVID-19 tests has been increased to quarantine infected people. However, due to the problems encountered in the supply of RTPCR tests and the ease of obtaining Computed Tomography and X-ray images, imaging-based methods have become very popular in the diagnosis of COVID-19. Therefore, studies using these images to classify COVID-19 have increased. This paper presents a classification method for computed tomography chest images in the COVID-19 Radiography Database using features extracted by popular Convolutional Neural Networks (CNN) models (AlexNet, ResNet18, ResNet50, Inceptionv3, Densenet201, Inceptionresnetv2, MobileNetv2, GoogleNet). The determination of hyperparameters of Machine Learning (ML) algorithms by Bayesian optimization, and ANN-based image segmentation are the two main contributions in this study. First of all, lung segmentation is performed automatically from the raw image with Artificial Neural Networks (ANNs). To ensure data diversity, data augmentation is applied to the COVID-19 classes, which are fewer than the other two classes. Then these images are applied as input to five different CNN models. The features extracted from each CNN model are given as input to four different ML algorithms, namely Support Vector Machine (SVM), k-Nearest Neighbors (k-NN), Naive Bayes (NB), and Decision Tree (DT) for classification. To achieve the most successful classification accuracy, the hyperparameters of each ML algorithm are determined using Bayesian optimization. With the classification made using these hyperparameters, the highest success is obtained as 96.29% with the DenseNet201 model and SVM algorithm. The Sensitivity, Precision, Specificity, MCC, and F1-Score metric values for this structure are 0.9642, 0.9642, 0.9812, 0.9641 and 0.9453, respectively. These results showed that ML methods with the most optimum hyperparameters can produce successful results.
  • Küçük Resim Yok
    Öğe
    A Novel Convolutional-Recurrent Hybrid Network for Sunn Pest-Damaged Wheat Grain Detection
    (Springer, 2022) Sabanci, Kadir; Aslan, Muhammet Fatih; Ropelewska, Ewa; Unlersen, Muhammed Fahri; Durdu, Akif
    The sunn pest-damaged (SPD) wheat grains negatively affect the flour quality and cause yield loss. This study focuses on the detection of SPD wheat grains using deep learning. With the created image acquisition mechanism, healthy and SPD wheat grains are displayed. Image preprocessing steps are applied to the captured raw images, then data augmentation is performed. The augmented image data is given as an input to two different deep learning architectures. In the first architecture, transfer learning application is made using AlexNet. The second architecture is a hybrid structure, obtained by adding the bidirectional long short-term memory (BiLSTM) layer to the first architecture. In terms of accuracy, the performance of the non-hybrid and hybrid architectures that are presented in the study is determined as 98.50% and 99.50%, respectively. High classification success and innovative deep learning structure are the features of this study that distinguish it from previous studies.
  • Küçük Resim Yok
    Öğe
    The Speed Estimation via BiLSTM-Based Network of a BLDC Motor Drive for Fan Applications
    (Springer Heidelberg, 2022) Unlersen, Muhammed Fahri; Balci, Selami; Aslan, Muhammet Fatih; Sabanci, Kadir
    In this study, in order to determine the dynamic response of a four-pole permanent magnet three-phase brushless DC (BLDC) motor, parametric simulation studies are carried out with finite element analysis Rmxprt software depending on three specific input variables (excitation voltage, pulse width, and motor power). The rotor speed is defined as the output parameter to determine the dynamic response, and 600 parametric data are obtained according to the simulation studies. In order to estimate the rotor speed of the BLDC motor modeled using artificial intelligence (AI), an advanced recurrent neural network architecture known as bidirectional long short-term memory has been designed. Rotor speed is successfully estimated with the proposed architecture, and as a result, the mean absolute percentage error value is calculated as 3.25%. These results show that the analysis of BLDC motor parameters can be determined quickly with the proposed AI method without long-running simulations.
  • Küçük Resim Yok
    Öğe
    Thermal behavior estimation of the power switches with an empirical formulation optimized by Artificial Bee Colony algorithm
    (Pergamon-Elsevier Science Ltd, 2021) Unlersen, Muhammed Fahri; Balci, Selami; Sabanci, Kadir
    Temperature rise and thermal management in power electronic circuits is a very important issue both electrically and in terms of more compact circuit design. Considering the non-linear temperature rise in the power switches, it is necessary to accurately estimate the temperature rise before the circuit can be experimentally executed. In this study, the temperature rise values in the switching elements for the different current, switching frequency, and duty ratio values of a DC-DC boost converter circuit are analyzed parametrically with ANSYS-Twin Builder software and a total of 715 data are obtained. Based on these data obtained, the temperature rise that occurs during the operation of the power switches in different parameters can be predicted and can be predicted accurately and quickly with a mathematical expression in order to effectively provide the cooling system and thermal management of the power electronics circuit. The proposed expression has eight parameters that need to be optimized. In optimization of these parameters, the Artificial Bee Colony algorithm is used. The Mean Absolute Error is used as the performance indicator. The temperature rise values are calculated via the optimized expression with 0.7446 degrees C error. Thus, the accuracy of expression is about 95.17%.