< previous page	page_135	next page >
Page 135 Table 4: P(x)and µx for the One-position Schemata Implicit in Table 3 x P(x) mx 1/3 0.9 1/3 1.1 1/3 1.0 1/3 1.1 1/3 1.0 1/3 0.9 Clearly the combination w2w1 becomes increasingly likely under ; in fact On the other hand, the best combination by the same calculation satisfies so that its probability of occurrence actually decreases. It is true that, as becomes more probable, the values of and decrease, eventually dropping below 1, but is still selected against, as the following table shows: Table 5: µx for the Schemata of Table 4 when Instances Are Not Equilikely x P(x) mx P(x) mx 0.01 0.620 0.01 0.066 0.90 0.982 0.09 1.080 0.09 1.270 0.90 0.947 0.90 0.982 0.09 1.080 0.09 1.270 0.90 0.947 0.01 0.620 0.01 0.066 Thus steadily decreases under , with a balance being struck among the schemata using weights {w2, w3} at position 1 and weights {w1, w2} at position 2. The lack of linkage between positions (or, equivalently, enforced operation at the equilibrium point l(x)) destroys the robustness of .
< previous page	page_135	next page >

If you like this book, buy it!