More recently, ability of Artificial Neural Network (ANN) in prediction of the production has led to increase in its applications in many fields of agriculture. Several researchers have shown that ANN has powerful pattern classification and pattern recognition capabilities. It has been applied for prediction of individual milk, fat and protein yields and prediction of some disease in dairy cattle. In the last decade, many studies have shown that ANN is relatively successful for prediction of production traits. The aim of this study was to evaluate the accuracy of prediction of 305 days milk production in dairy cattle by bpANN model (back propagation ANN). The designed network had 3 layers including input, hidden and output with 11, 50 and 1 neuron(s), respectively. From all of 100 records of data 75%, 12.5% and 12.5% was assigned to training, verification and test category, respectively. Test-day records were taken at 28-day intervals. In this study 5 test-day period records were used (from TD1 to TD5) and each test-day record contained information about previous test-day records. Accuracy of prediction varied from TD1= 0.77 to TD5= 0.85 and was the highest when 5 test-day records were used. The results indicated that illustrated ANN can readily be adapted with any new data that added to the network and solve complex problems. Also, the results demonstrated our ANN model has the ability to predict 305 days milk yields with test-day records. In addition using more test-day records can increase the ability of bpANN for prediction of 305 days milk production in dairy cattle. Abstracts - Abstract: preview Page 1 of 1 http://

Keywords