Improving Prediction Performance Using Ensemble Neural Networks in Textile Sector

Yildirim P., Birant D., Alpyıldız T.

2017 International Conference on Computer Science and Engineering (UBMK), Antalya, Türkiye, 5 - 08 Ekim 2017, ss.639-644 identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası:
  • Doi Numarası: 10.1109/ubmk.2017.8093487
  • Basıldığı Şehir: Antalya
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.639-644
  • Anahtar Kelimeler: prediction, ensemble learning, neural network, multilayer perceptron
  • Dokuz Eylül Üniversitesi Adresli: Evet

Özet

Neural network technique has been recently preferred in textile sector for the prediction task because the traditional mathematical and statistical methods can be inadequate to derive complex relations within textile datasets. Meanwhile ensemble learning has become a popular machine learning approach in recent years due to the high prediction performance it provides. Therefore, this study proposes an ensemble learning approach that combines neural networks with different parameter values (the number of hidden layers, learning rate and momentum coefficient) to improve prediction performance in textile sector. In the experimental studies, the proposed model was tested on ten different real-world textile datasets. The results show that ensemble neural networks usually achieve better prediction performance than an individual neural network in terms of correlation coefficient and relative absolute error measures.