A Hybrid Approach for Improving the Classification Performance


KÖK İ., Davarci M. E., Ozdemir S.

2017 International Conference on Computer Science and Engineering (UBMK), Antalya, Türkiye, 5 - 08 Ekim 2017, ss.682-687 identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası:
  • Doi Numarası: 10.1109/ubmk.2017.8093498
  • Basıldığı Şehir: Antalya
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.682-687
  • Gazi Üniversitesi Adresli: Evet

Özet

There are many factors that affect the performance of classification. The volume, size, type of data and classification methods are the most obvious factors. For the exact same data set, it is possible to achieve different classification performance values by using different classification methods Hence, the development of classification models that arc more accurate and applicable to many areas for classification problems has a great importance. In this work, a hybrid classification model combining Naive Bayes, Perceptron and KNN is proposed. In this model, a novel parameter called Decision Function is used. The proposed decision function aims to increase the classification success by considering the classification results of the three algorithms The performance evaluation results show that the proposed decision function significantly improves the classification success.