Information Technology Reference
In-Depth Information
As a matter of fact, the greatest challenge in time series prediction con-
sists of avoiding these strong attractors in the solution landscape. And this is
no doubt one of the reasons why Kolmogorov-Gabor polynomials are par-
ticularly suited for this task, for it is almost impossible to stumble upon these
points in the solution space with them. Notwithstanding, as we will see in the
next section, Kolmogorov-Gabor polynomials are much less efficient than
the much simpler GEP systems. Furthermore, we will also see that these
polynomials have another great disadvantage: they are immensely complex,
making it almost impossible to extract knowledge from them.
Just for the sake of curiosity, let's take a look at the structure of the best-
of-experiment solutions designed with Kolmogorov-Gabor polynomials. The
best solution of the GEP-OS experiment was discovered in generation 4927
of run 45. Its structure is shown below:
F9.F9.F9.j.F9.e.d.b.i.a.j.F9.a.F9.F9.g.g.F9.F9.F9.F9.g.d.e.j.d.g.c.b.
b.e.d.j.e.b.h.e.c.g.b.b.h.e.r112.r58.r105.r18.r18.r18.r6.r66.r13.r18.
r41.r41.r13.r12.r109.r64.r63.r63.r80.r21.r85.r87.r87.r87.r103.r18.
r10.r100.r83.r89.r51.r42.r14.r106.r58.r68.r92.r3.r67.r51.r87.r103.
r20.r105.r8.r9.r98.r72.r17.r72.r99.r84.r29.r55.r40.r47.r56.r1.r116.
r80.r80.r45.r79.r88.r24.r98.r45.r51.r21.r38.r32.r44.r40.r24.r43.
r99.r12.r93.r11.r118.r33.r14.r57.r107.r36.r5.r62.r12.r55.r117.r50.
r87.r116.r76.r100.r72.r117.r29.r52.r61.r13.r108.r8.r42.r103.r21.
r32.r64.r62.r103.r109.r91.r3.r16.r30.r7.r92.r61.r90.r35.r99.r61.
r12.r78.r0.r110
C = {-0.771332, 0.042266, -0.443359, -0.478027, -0.365906,
0.437317, -0.749542, -0.30246, -0.751587, 0.050018,
-0.300842, 0.646515, -0.932831, -0.369415, -0.692658,
0.234894, -0.192352, -0.707092, -0.003479, -0.703156,
-0.570465, -0.995209, -0.845459, -0.756562, -0.072815,
-0.67981, -0.274902, -0.528076, 0.051422, -0.221344,
-0.573303, -0.833222, 0.280884, -0.180175, -0.343201,
0.426483, 0.355102, -0.797303, 0.382172, -0.835694,
0.043457, -0.002349, -0.961548, 0.287567, 0.840149,
-0.263611, 0.44754, -0.083435, 0.153839, -0.817536,
0.270538, -0.256012, -0.587372, 0.469818, 0.378082,
0.09375, 0.302521, -0.56604, 0.847077, -0.699921,
-0.080535, 0.599335, -0.084106, 0.04776, -0.093383,
-0.561371, -0.995942, -0.322906, -0.039794, 0.830567,
-0.999817, -0.671631, -0.217529, 0.15271, 0.84372,
