종모우모형에서 직접법과 반복법의 해의 비교 = Comparison of Solutions from G- inverse and Iterative Method with Sire Model

다국어 초록 (Multilingual Abstract)

For the evaluation of breeding values of animals, A trial to get solutions from the Iterative(SUR: Successive underrelaxation method) and the G-inverse was carried out using a sire model with 76 levels of 7 independent factors in this study. The results showed no differences between the solutions from directive(G-inverse) method and that from iterative method. For the four(0.1, 0.2, 0.3, 0.35) different relaxation factors(W) to speed up the convergence to the final solution, even though the number of rounds of interaction was affected, there were no differences in the final solutions. The numbers of rounds of iteration t get final solutions from a restriction({ b₁^n+1 - b₁} $lt;0.0001) under for the 0.1, 0.2, 0.3 and 035 relaxation factors were 1729. 928, 649 and 563 rounds respectively. It is the good iteration parameter(w) whether the convergence is speeded up or not. Since Iterative method is very simple to be programized and needs relatively small computer memory, it could be useful for solving large and complex equations in animal evaluation.