Scopus İndeksli Yayınlar Koleksiyonu

Permanent URI for this collectionhttps://hdl.handle.net/20.500.12416/8651

Browse

Filters

2013 - 201932020 - 20255

Settings

Subject: search.filters.subject.Feature Extraction

Search Results

Now showing 1 - 8 of 8
  • Conference Object
    Deep Learning Model for Fingerprint Biometric Identification System
    (Institute of Electrical and Electronics Engineers Inc., 2025) Abdulkarim, Anas; Ulu, Eren; Sever, Hayri
  • Article
    Detection and Classification of Femoral Neck Fractures From Plain Pelvic X-Rays Using Deep Learning and Machine Learning Methods
    (Turkish Assoc Trauma Emergency Surgery, 2025) Sevinc, Huseyin Fatih; Ureten, Kemal; Karadeniz, Talha; Gultekin, Gokhan Koray
    Background: Femoral neck fractures are a serious health concern, particularly among the elderly. The aim of this study is to diagnose and classify femoral neck fractures from plain pelvic X-rays using deep learning and machine learning algorithms, and to compare the performance of these methods. Methods: The study was conducted on a total of 598 plain pelvic X-ray images, including 296 patients with femoral neck fractures and 302 individuals without femoral neck fractures. Initially, transfer learning was applied using pre-trained deep learning models: VGG-16, ResNet-50, and MobileNetv2. Results: The pre-trained VGG-16 network demonstrated slightly better performance than ResNet-50 and MobileNetV2 for detecting and classifying femoral neck fractures. Using the VGG-16 model, the following results were obtained: 95.6% accuracy, 95.5% sensitivity, 93.3% specificity, 95.7% precision, 95.5% F1 Score, a Cohen's kappa of 0.91, and the Receiver Operating Characteristic (ROC) curve of 0.99. Subsequently, features extracted from the convolution layers of VGG-16 were classified using common machine learning algorithms. Among these, the k-nearest neighbor (k-NN) algorithm outperformed the others and exceeded the accuracy of the VGG-16 model by 1%. Conclusion: Successful results were obtained using deep learning and machine learning methods for the detection and classification of femoral neck fractures. The model can be further improved through multi-center studies. The proposed model may be especially useful for physicians working in emergency departments and for those not having sufficient experience in evaluating plain pelvic radiographs.
  • Conference Object
    Citation - Scopus: 1
    Localization of Semantic Category Classification in Fmri Images
    (Ieee, 2014) Alkan, Sarper; Yarman-Vural, Fatos T.
    In this study, we provide a methodology to localize the brain regions that contribute to semantic category classification. For this purpose we first cluster the data using spectral clustering. Then we extract local features within each cluster by using mesh-arc descriptors. Finally, we test the classification accuracy of each cluster against a hypothesis testing measure we provide here. We have found that, for the experimental task at hand, calcerine fissure and angular gyrus were most effective in classification. These results are shown to be match well with the nature of the experiment. Thus the validity of our approach is confirmed.
  • Conference Object
    Spam Detection With Fasttext Based Features
    (Institute of Electrical and Electronics Engineers Inc., 2024) Karadeniz, T.; Tokdemir, G.; Maraş, H.H.
    Fasttext is a powerful word representation method that creates word representations based on vectors of character n-grams. In this work, we propose a method that utilizes fasttext features for a novel feature engineering model for the spam detection problem. In the feature engineering method, the combination of average, mean of second derivative; mean peak and standard deviation of fasttext features are computed. Finally, tf-idf features are also considered for the modeling process. The success of each feature engineering technique is measured and reported. The combination of the five feature extraction methods, tested on two spam detection datasets, yielded promising results with an accuracy of 0.978 on e-mail spam detection and an accuracy of 0.986 on sms spam classification. © 2024 IEEE.
  • Article
    Citation - Scopus: 7
    Robust Classification for Sub Brain Tumors by Using an Ant Colony Algorithm With a Neural Network
    (Innovative Information Science and Technology Research Group, 2024) Faris, R.A.; Mosa, Q.; Albdairi, M.
    A brain tumor is responsible for the highest number of fatalities across the globe. Identifying and diagnosing the tumor correctly at an early stage can significantly improve the chances of survival. Classifying a brain tumor can be aided by factors like type, texture, and location. In this research, we propose a robust technique for detecting sub-brain tumors using an ant colony algorithm coupled with a neural network. To achieve this, we employ an ant colony optimization algorithm (ACO) to eliminate extraneous features extracted from the image, enabling us to find the most effective representation of the image. This, in turn, assists the Neural Network (NN) in the process of classification. Our system involves a series of five steps. Initially, we perform cropping processing as the initial step to eliminate unnecessary background from the original MRI images. This enhances the overall quality of the images, thereby improving the performance of the classification method. In the next step, we conduct image preprocessing to enhance image quality, making it easier for the feature extractor to accurately extract features. The third step involves employing a feature extractor for each image. In the fourth step, we utilize the ant colony optimization algorithm (ACO) to identify the most suitable representation of the image, which further aids the NN in classification. In the fifth and final step, we utilize an NN method to classify the vector obtained from the fourth step (optimization method) to determine the subtype of the brain tumor (normal, glioma, meningioma, and pituitary). Our model's performance is evaluated using the publicly available BT-large-4c dataset, and it surpasses current state-of-the-art methods with exceptional accuracy, attaining a rate of 87.7%. The effectiveness of our approach is particularly evident in maintaining accurate classifications within MRI input images. © 2024, Innovative Information Science and Technology Research Group. All rights reserved.
  • Article
    Citation - WoS: 10
    Citation - Scopus: 14
    A Pairwise Deep Ranking Model for Relative Assessment of Parkinson's Disease Patients From Gait Signals
    (Ieee-inst Electrical Electronics Engineers inc, 2022) Ogul, Burcin Buket; Ozdemir, Suat
    Continuous monitoring of the symptoms is crucial to improve the quality of life for patients with Parkinson's Disease (PD). Thus, it is necessary to objectively assess the PD symptoms. Since manual assessment is subjective and prone to misinterpretation, computer-aided methods that use sensory measurements have recently been used to make objective PD assessment. Current methods follow an absolute assessment strategy, where the symptoms are classified into known categories or quantified with exact values. These methods are usually difficult to generalize and considered to be unreliable in practice. In this paper, we formulate the PD assessment problem as a relative assessment of one patient compared to another. For this assessment, we propose a new approach to the comparative analysis of gait signals obtained via foot-worn sensors. We introduce a novel pairwise deep-ranking model that is fed by data from a pair of patients, where the data is obtained from multiple ground reaction force sensors. The proposed model, called Ranking by Siamese Recurrent Network with Attention, takes two multivariate time-series as inputs and produces a probability of the first signal having a higher continuous attribute than the second one. In ten-fold cross-validation, the accuracy of pairwise ranking predictions can reach up to 82% with an AUROC of 0.89. The model outperforms the previous methods for PD monitoring when run in the same experimental setup. To the best of our knowledge, this is the first study that attempts to relatively assess PD patients using a pairwise ranking measure on sensory data. The model can serve as a complementary model to computer-aided prognosis tools by monitoring the progress of the patient during the applied treatment.
  • Conference Object
    Citation - Scopus: 1
    An Isolated Word Speaker Recognition System
    (Institute of Electrical and Electronics Engineers Inc., 2017) Ozaydin, S.
    The paper presents a design of an isolated word speaker recognizer system by using the Hidden Markov Model. HTK toolkit is used as a design tool. The system is operated on a closed set of speakers and with a limited vocabulary of words. Digit utterances from one to ten with ten speakers (7 male, 3 female) are used as dataset in the system. It consists of isolated words that are separated by silences. Each speaker reads each word ten times. Six of them are used in training and the remaining are used in the test phase. The Mel cepstral coefficients are used in order to design the acoustic features and HMM models are constructed. A threshold calculation is performed for each speaker. After threshold adjustment, tests are performed for the performance evaluation and accuracy rates are calculated for each user. The system resulted in changing levels of recognition accuracy for each speaker and each word. © 2017 IEEE.
  • Article
    Citation - WoS: 35
    Citation - Scopus: 46
    Fall Detection Using Single-Tree Complex Wavelet Transform
    (Elsevier, 2013) Keskin, Furkan; Toreyin, B. Ugur; Cetin, A. Enis; Yazar, Ahmet
    The goal of Ambient Assisted Living (AAL) research is to improve the quality of life of the elderly and handicapped people and help them maintain an independent lifestyle with the use of sensors, signal processing and telecommunications infrastructure. Unusual human activity detection such as fall detection has important applications. In this paper, a fall detection algorithm for a low cost AAL system using vibration and passive infrared (PIR) sensors is proposed. The single-tree complex wavelet transform (ST-CWT) is used for feature extraction from vibration sensor signal. The proposed feature extraction scheme is compared to discrete Fourier transform and mel-frequency cepstrum coefficients based feature extraction methods. Vibration signal features are classified into "fall" and "ordinary activity" classes using Euclidean distance, Mahalanobis distance, and support vector machine (SVM) classifiers, and they are compared to each other. The PIR sensor is used for the detection of a moving person in a region of interest. The proposed system works in real-time on a standard personal computer. (C) 2012 Elsevier B.V. All rights reserved.