Browsing by Author "Koçyiǧit Y."

Now showing 1 - 6 of 6
  • No Thumbnail Available
    Item
    Comparison of signal processing methods used for classification of EMG signals; [EMG işaretlerini siniflamada kullanilan işaret işleme tekniklerinin karşilaştirilmasi]
    (2005) Karlik B.; Koçyiǧit Y.; Fidan C.B.
    The Electromyographic (EMG) signals observed at the surface of the skin is the sum of many small action potentials generated in the muscle fibers. There is only a pattern for each EMG signals, which are generated by biceps and triceps muscles. This pattern consists of information of direction and speed of movement. There are different types of signal processing to find feature values for true classification in this pattern. In this study, the Feature values belong to 4 different arm movements are obtained by using Autoregressive parameters (AR), Fast Fourier Transform (FFT), and Wavelet Transform. Then these feature values are compared each other by using same classifier. The Back-Propagation Algorithm, which has 3 layer perception, was preferred as a classifier. © 2005 IEEE.
  • No Thumbnail Available
    Item
    Using LBG algorithm for extracting the features of EMG signals; [EMG i̇şaretlerinin özniteliklerinin belirlenmesinde LBG algoritmasinin kullanimi]
    (2008) Koçyiǧit Y.; Kiliç I.
    The Electromyographic (EMG) signals observed at the surface of the skin is the sum of many small action potentials generated in the muscle fibers. There is only a pattern for each EMG signals, which are generated by biceps and triceps muscles. There are different types of signal processing in order to find out the feature values for true classification in this pattern. In this study, the Feature values belong to 4 different arm movements are obtained by using clustering methods, i.e K-means, Fuzzy C-means, and LBG after applying Wavelet Transform to EMG signals . Then these feature values are compared each other by KEYK and Quadratic Discriminant Analysis classifier. ©2008 IEEE.
  • No Thumbnail Available
    Item
    Determination of chemisorption probabilities of hydrogen molecules on a nickel surface by artificial neural network
    (2008) Böyükata M.; Koçyiǧit Y.; Güvenç Z.B.
    Dissociative chemisorption probabilities for H2(v, j) + Ni(100) collision systems have been estimated by using Artificial Neural Network (ANN). For training, previously determined probability values via molecular dynamics simulations have been used. Performance of the ANN, for predicting any quantities in the molecule-surface interaction, has been investigated. Effects of the surface sites and the rovibrational states of the molecule on the process are analyzed. The results are in good agreement with the related previous studies.
  • No Thumbnail Available
    Item
    Fast global fuzzy C-means clustering for ECG signal classification; [EKG i̇şaretlerini siniflamak için hizli global bulanik C-ortalama öbekleşme]
    (2010) Koçyiǧit Y.; Kiliç I.
    Fuzzy clustering plays an important role in solving problems in the areas of pattern recognition and fuzzy model identification. The Fuzzy C-Means algorithm is one of widely used algorithms. It is based on optimizing an objective function, being responsive to initial conditions; the algorithm usually leads to local minimum results. Aiming at above problem, the fast global Fuzzy C-Means clustering algorithm (FGFCM) has been proposed, which is an incremental approach to clustering, and does not depend on any initial conditions. The algorithm was applied on ECG signals to classification. ©2010 IEEE.
  • No Thumbnail Available
    Item
    Heart sound signal classification using fast independent component analysis
    (Turkiye Klinikleri Journal of Medical Sciences, 2016) Koçyiǧit Y.
    The analysis of heart sound signals is a basic method for heart examination. It may indicate the presence of heart disorders and provide clinical information in the diagnostic process. In this study, a novel feature dimension reduction method based on independent component analysis (ICA) has been proposed for the classification of fourteen different heart sound types; the method was compared with principal component analysis. The feature vectors are classified by support vector machines, linear discriminant analysis, and naive Bayes (NB) classifiers using 10-fold cross validation. The ICA combined with NB achieves the highest average performance with a sensitivity of 98.53%, specificity of 99.89%, g-means of 99.21%, and accuracy of 99.79%. © 2016 Tübitak.
  • No Thumbnail Available
    Item
    A new approach to genetic algorithm in image compression
    (Institute of Electrical and Electronics Engineers Inc., 2017) Harman F.; Koçyiǧit Y.
    The importance of image compression problem has been progressing with the development of technology. The usage of genetic algorithm has become widespread in this field. In this study, the general structure of genetic algorithm and its effects on image compression are analyzed. In this study, it is seen that the creation of population via natural selection, the ratio of mutation and crossover affect the performance of image compression a lot. Roulette Wheel Selection and Elitist Selection that are the most known natural selections are firstly implemented on the standard image. But with these known natural selections, MSE (mean square error) and PSNR (peak signal noise ratio) are seen close to each other. It is seen that in all implementation with the 10% crossover and 5% mutation ratio, the natural selection algorithm based on pools has better MSE and PSNR values than genetic algorithm based on roulette wheel and elitist selection respectively. © 2017 EMO (Turkish Chamber of Electrical Enginners).