212 VLSI Test Principles and Architectures TABLE 4.9 ± Uniform Crossover Mask 010011100100010011110101 Parent #1 110011001100110011001100 Parent #2 101010101010101010101010 Child #1 100010101000100010101000 Child #2 111011001110111011001110 TABLE 4.10 ± Mutating Bit Position #3 Before mutation 110011001100110011001100 After mutation 111011001100110011001100 The third GA operator to be discussed is the mutation operator. It allows the child individual to vary slightly from the two parents it had inherited. The mutation operator simply selects a random bit position in an individual and flips its logic value with a mutation probability. An example of mutating bit #3 is shown in Table 4.10. Let ± be the mutation probability. If ± is too small, children individuals that are produced after crossover may rarely see any mutation. In other words, it is less likely that new genes (building blocks) will be produced. On the other hand, if ± is too large, too much random perturbation may occur, and the resemblance
This is the end of the preview. Sign up
access the rest of the document.