Yazar "Gunes, S" seçeneğine göre listele

Listeleniyor 1 - 5 / 5
  • Küçük Resim
    Öğe
    Artificial immune systems and ANN in PMR coding
    (IEEE, 2004) Sahan, S; Ceylan, R; Gunes, S
    This study aims at introducing a new Artificial Intelligence technique especially beginned to be known newly in our country and models human immune system. Artificial Immune Systems (AIS) are increasing their application areas day by day and coming into existence as a prosperius problem solving technique with their performance on these applications. ABNET is an hybrid system proposed by F. N. De Castro, F. J. Von Zuben and Getulio A. De Deus Jr. that is a combination of Artificial Neural Networks and immune based metaphors. In this study, the coding in PMR (Private Mobile Radio) is performed with ABNET. To evaluate the results, same coding problem is solved with an ANN system, too. With respect to the results of both system, the applicability of ABNET to real world problems is discussed.
  • Küçük Resim
    Öğe
    The medical applications of attribute weighted artificial immune system (AWAIS): Diagnosis of Heart and Diabetes Diseases
    (SPRINGER-VERLAG BERLIN, 2005) Sahan, S; Polat, K; Kodaz, H; Gunes, S
    In our previous work, we had been proposed a new artificial immune system named as Attribute Weighted Artificial Immune System (AWAIS) to eliminate the negative effects of taking into account of all attributes in calculating Euclidean distance in shape-space representation which is used in many network-based Artificial Immune Systems (AISs), This system depends on the weighting attributes with respect to their importance degrees in class discrimination. These weights are then used in calculation of Euclidean distances. The performance analyses were conducted in the previous study by using machine learning benchmark datasets. In this study, the performance of AWAIS was investigated for real world problems. The used datasets were medical datasets consisting of Statlog Heart Disease and Pima Indian Diabetes datasets taken from University of California at Irvine (UCI) Machine Learning Repository. Classification accuracies for these datasets were obtained through using 10-fold cross validation method. AWAIS reached 82.59% classification accuracy for Statlog Heart Disease while it obtained a classification accuracy of 75.87% for Pima Indians Diabetes. These results are comparable with other classifiers and give promising performance to AWAIS for that kind of problems.
  • Küçük Resim
    Öğe
    A new classification method for breast cancer diagnosis: Feature Selection Artificial Immune Recognition System (FS-AIRS)
    (SPRINGER-VERLAG BERLIN, 2005) Polat, K; Sahan, S; Kodaz, H; Gunes, S
    In this study, diagnosis of breast cancer, the second type of the most widespread cancer in women, was performed with a new approach, FS-AIRS (Feature Selection Artificial Immune Recognition System) algorithm that has an important place in classification systems and was developed depending on the Artificial Immune Systems. With this purpose, 683 data in the Wisconsin breast cancer dataset (WBCD) was used. In this study, differently from the studies in the literature related to this concept, firstly, the feature number of each data was reduced to 6 from 9 in the feature selection sub-program by means of forming rules related to the breast cancer data with the C4.5 decision tree algorithm. After separating the 683 data set with reduced feature number into training and test sets by 10 fold cross validation method in the second stage, the data set was classified in the third stage with AIRS and a quite satisfying result was obtained with respect to the classification accuracy compared to the other methods used for this classification problem.
  • Küçük Resim
    Öğe
    A new classifier based on attribute weighted artificial immune system (AWAIS)
    (SPRINGER-VERLAG BERLIN, 2004) Sahan, S; Kodaz, H; Gunes, S; Polat, K
    'Curse of Dimensionality' problem in shape-space representation which is used in many network-based Artificial Immune Systems (AISs) affects classification performance at a high degree. In this paper, to increase classification accuracy, it is aimed to minimize the effect of this problem by developing an Attribute Weighted Artificial Immune System (AWAIS). To evaluate the performance of proposed system, aiNet, an algorithm that have a considerably important place among network-based AIS algorithms, was used for comparison with our developed algorithm. Two artificial data sets used in aiNet, Two-spirals data set and Chainlink data set were applied in the performance analyses, which led the results of classification performance by means of represented network units to be higher than aiNet. Furthermore, to evaluate performance of the algorithm in a real world application, wine data set that taken from UCI Machine Learning Repository is used. For the artificial data sets, proposed system reached 100% classification accuracy with only a few numbers of network units and for the real world data set, wine data set, the algorithm obtained 98.23% classification accuracy which is very satisfying result if it is considered that the maximum classification accuracy obtained with other systems is 98.9%.
  • Küçük Resim
    Öğe
    Outdoor image classification using artificial immune recognition system (AIRS) with performance evaluation by fuzzy resource allocation mechanism
    (SPRINGER-VERLAG BERLIN, 2005) Polat, K; Sahan, S; Kodaz, H; Gunes, S
    AIRS classification algorithm, which has an important place among classification algorithms in the field of Artificial Immune Systems, has showed an effective and intriguing performance on the problems it was applied. In this study, the resource allocation mechanism of AIRS was changed with a new one determined by Fuzzy-Logic rules. This system, named as Fuzzy-AIRS and AIRS were used as classifiers in the classification of outdoor images. The classification of outdoor dataset taken from UCI repository of machine learning databases was done using 10-fold cross validation method. Both versions of AIRS well performed over other systems reported in UCI website for corresponding dataset. Fuzzy-AIRS reached to the classification accuracy of 90.00 % in the applications whereas AIRS obtained 88.20 %. Besides, Fuzzy-AIRS gained one more advantage over AIRS by means of classification time. In the experiments, it was seen that the classification time in Fuzzy-AIRS was reduced by about 67% of AIRS for dataset. Fuzzy-AIRS classifier proved that it can be used as an effective classifier for image classification by reducing classification time as well as obtaining high classification accuracies.