Information Technology Reference
In-Depth Information
functions on its own by choosing a cellular system. In the analysis shown in
Figure 4.5, multiplication was used to link the sub-ETs. The remaining pa-
rameters are exactly the same as in the analysis shown in Figure 4.3.
As expected, for this polynomial function, the algorithm excelled with
multigenic chromosomes encoding sub-ETs posttranslationally linked by
addition. In real-world problems this testing of waters is done until a feel for
the best chromosomal structure and composition is developed, and this usu-
ally entails making 3 or 4 preparatory runs, not the exhaustive analysis done
here. Typically, one usually experiments with a couple of chromosomal or-
ganizations and tests different function sets. By observing such indicators as
best and average fitnesses, it is easy to see whether the system is evolving
efficiently or not. Then, after choosing the appropriate settings, one just lets
the system evolve the best possible solution on its own.
Consider, for instance, the multigenic system composed of four genes linked
by addition. As shown in Figure 4.3, the success rate has, in this case, the
maximum value of 100%, meaning that, in all the runs, a perfect solution
was found. Let's analyze with more detail a successful run of this experi-
ment. The parameters used per run are summarized in Table 4.1 above.
100
90
80
70
60
50
40
30
20
10
0
0
1
2
3
4
5
6
7
8
Number of genes
Figure 4.5. Variation of success rate with the number of genes. For this analysis,
G = 50, P = 30, and h = 6 (corresponding to a gene size of 13). The sub-ETs were
linked by multiplication. The success rate was evaluated over 100 identical runs.
 
Search WWH ::




Custom Search