| Yazarlar |
Doç. Dr. Ramazan YAZICI
Kırşehir Ahi Evran Üniversitesi, Türkiye |
| Özet |
| Determination of the condition factor in fish is an indispensable element in protecting fish health and improving the status of the population. In this study, the condition factor (CF) of fish was predicted using three input parameters including length, weight and sex. In this paper, the results obtained with six machine learning algorithms; Support vector machine (SVM), Neural Network/Multilayer Perceptron (MLP), Ensemble Learning, Gaussian Process Regression (GSR), Decision Tree and Linear Regression were compared with a multilayer perceptron artificial neural network (MLP-ANN), which is one of the statistical tools to predict the condition factor value obtained in this paper. As a result of the benchmarking, the Levenberg-Marquardt learning algorithm with 3-9-1 architecture neurons was found to be the best network for the hidden layer. The output of this model was the most effective for condition factor modeling with R 2 values (R 2= training (1), testing (0.99), validation (1) and overall (0.99)). This value is indicative of the high quality of this model compared to other existing models. Up to now, multilayer perceptron artificial neural network (MLP-ANN) has achieved significant success in predicting the condition factor. |
| Anahtar Kelimeler |
| Makale Türü | Özgün Makale |
| Makale Alt Türü | Uluslararası alan indekslerindeki dergilerde yayımlanan tam makale |
| Dergi Adı | Acta Biologica Turcica |
| Dergi Tarandığı Indeksler | |
| Makale Dili | İngilizce |
| Basım Tarihi | 01-2024 |
| Cilt No | 37 |
| Sayı | 1 |
| Sayfalar | 8 / 1 |
