DETERMINING THE BEST NORMALIZATION TECHNIQUE FOR ESTIMATION USING ARTIFICIAL NEURAL NETWORKS: CASE OF BRUSHTOOTH LIZARDFISH

SANGÜN L.

FRESENIUS ENVIRONMENTAL BULLETIN, cilt.28, sa.4, ss.2842-2847, 2019 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 28 Sayı: 4
  • Basım Tarihi: 2019
  • Dergi Adı: FRESENIUS ENVIRONMENTAL BULLETIN
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED)
  • Sayfa Sayıları: ss.2842-2847
  • Anahtar Kelimeler: Artificial Neural Networks, normalization, estimation, Brushtooth lizardfish, MULTIPLE LINEAR-REGRESSION, SAURIDA-UNDOSQUAMIS, ALGORITHM, SYSTEM
  • Çukurova Üniversitesi Adresli: Evet

Özet

In this study, the bodyweight of Brushtooth lizardfish was estimated through the use of artificial neural networks (ANNs) by applying various normalization techniques to the morphometric data (total length, fork length, standard length) of the fish, and the best normalization method was selected based on the results. Z-Score, Median, Sigmoid, Min-max and D-Min-Max methods were applied in the given order, and the best MAPE and MSE values in the ANNs were calculated to be 3.187-0.001 for D-Min-Max and 3.784-0.001 for Min-max. Since the estimates obtained from the application of these two methods will turn out to be more accurate according to the results of ANN analyses, they are the methods recommended to be employed.