Abstract

Concentration of semen production is the most important fertility trait in ram and dimension of testis is a good criterion for identifying the quantity of semen production. Thus, prediction of that trait has important beneficial effect on the timely identification of genetically superior animals. Artificial neural network (ANN) system can be used as a decision making support system in ram industry as well as other industries. It can help breeders to predict future semen production based on phenotypic trait. Data from 24 rams of zandi breed in Tehran, Iran, were used. From 192 available data of phenotypic and semen concentration, 184 records were used for training a back propagation ANN system and 8 randomly chosen record (not used in the training process) were introduced to the trained neural network for evaluation. The result of the simulation showed that there was no significant difference between the observed and the predicted semen production (p > 0.05). The major use of this predictive system is to make accurate selection decision which is based on prior knowledge of the outcomes.

Key words: Artificial neural network, correlation, semen production, ram.

Abbreviation

SP, Semen production; ANN, artificial neural network; RMSE, root mean square error.