IEEE CEC Submisson Aleshunas Janikow (7 Apr 11) v1 2
IEEE CEC Submisson Aleshunas Janikow (7 Apr 11) v1 2
IEEE CEC Submisson Aleshunas Janikow (7 Apr 11) v1 2
Abstract—Constrained Genetic Programming (CGP) is a evaluations. In crossover, randomly chosen sub-trees are
method of searching the Genetic Programming search space exchanged between two parents. In mutation, a randomly
non-uniformly, giving preferences to certain subspaces chosen sub-tree is regrown.
according to some heuristics. Adaptable CGP (ACGP) is a
method for discovery of the heuristics. CGP and ACGP have
previously demonstrated their capabilities using first-order
heuristics: parent-child probabilities. Recently, the same
advantage has been shown for second-order heuristics:
parent-children probabilities. A natural question to ask is
whether we can benefit from extending ACGP with deeper-
order heuristics. This paper attempts to answer this question
by performing cost-benefit analysis while simulating the
higher-order heuristics environment. We show that this
method cannot be extended beyond the current second or
possibly third-order heuristics without a new method to deal
with the sheer number of such deeper-order heuristics.
implicitly, but when derived from first order heuristics all 0.9
0.7
1 0.5
2d OH
0.4 1st OH
0.9
Base
0.8 0.3
0.7 0.2
0.6 0.1
Fitness
0.5 0
100
110
120
130
140
150
160
170
0
10
20
30
40
50
60
70
80
90
0.4 Seconds
0.3
0.2 2d OH
1st OH
Figure 4 Comparison of Base, 1st Order and 2d Order Heuristics
0.1
Base for Equation 3 180 seconds, 500 population
0
0
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
270
280
290
300
310
320
330
340
350
360
370
380
390
400
410
420
430
440
450
460
470
480
490
500
20.00 19.20
18.00 10.000
12.00 1 2 3 4 5 6 7 8 9 10 11
Seconds
Concurrent Processes
10.00
7.93
8.00