APPENDIX
B – Sample code
output
SAMPLE
OUTPUT FROM TEST WW1-MODBREEDER-6
AppAccelerator(tm)
1.1.034 for Java (JDK 1.1), x86 version.
Copyright
(c) 1998 Borland International. All Rights Reserved.
Welcome
to the network breeder (type 'help' for a list of commands).
->
modbreeder
modbreeder
- set up a modular network breeder, and breed modular networks.
please
enter following parameters (space seperated);
number
of inputs: 64
number
of outputs: 3
number
of modular networks: 5
initial
number of modules in each network: 3
maximun
number of layers in randomly generated networks: 2
maximum
number of nodes per layer in randomly generated networks: 5
number
of training patterns: 50
filename
of training inputs data: h:\uni\project\data\what-where\ww1-inputs
filename
of training outputs data: h:\uni\project\data\what-where\ww1-outputs
number
of testing patterns: 100
filename
of testing inputs data: h:\uni\project\data\what-where\ww1-test-inputs
filename
of testing outputs data: h:\uni\project\data\what-where\ww1-test-outputs
number
of epochs to train modular networks: 50
number
of epochs to train each population member: 50
desired
error: 0.1
percentage
improvement that breeder must achieve: 30
module
training step constant: 0.1
modular
network training memory constant: 0.1
modular
network training step constant: 0.1
modular
network training improvement constant: 1.0
percentage
of population to replace: 50
percentage
of population to mutate: 50
monitoring
on? (y/n): y
filename
for cycle monitoring:
h:\uni\project\monitor\ww1-modbreeder-6\ww1-modbreeder-cycles-6
IO
exception, probably incorrect filename
filename
for cycle monitoring:
h:\uni\project\monitors\ww1-modbreeder-6\ww1-modbreeder-cycles-6
filename
for error monitoring:
h:\uni\project\monitors\ww1-modbreeder-6\ww1-modbreeder-errors-6
evolving
modular networks to create best network...
setting
up initial population of modular networks
training
all modular networks
TEST:
training all modules and gating network, on all examples (for 1 epoch per
example),
for 50 epochs
m...............................................................................
-EDITED-
..............................
TEST:
average error per epoch of module 1 = 0.41409505383830586
TEST:
average weight per epoch of module 1 = 0.05723214476484701
TEST:
average error per epoch of module 2 = 0.1977500647031548
TEST:
average weight per epoch of module 2 = 0.48765896674784465
TEST:
average error per epoch of module 3 = 0.2283812008382145
TEST:
average weight per epoch of module 3 = 0.45538084793551853
TEST:
training all modules and gating network, on all examples (for 1 epoch per
example),
for 50 epochs
m...............................................................................
-EDITED-
..............................
TEST:
average error per epoch of module 1 = 0.3593454122047755
TEST:
average weight per epoch of module 1 = 0.21801218370762906
TEST:
average error per epoch of module 2 = 0.3948502128677635
TEST:
average weight per epoch of module 2 = 0.21695961178077497
TEST:
average error per epoch of module 3 = 0.2591488590363779
TEST:
average weight per epoch of module 3 = 0.27257606758522274
TEST:
training all modules and gating network, on all examples (for 1 epoch per
example),
for 50 epochs
m...............................................................................
-EDITED-
..............................
TEST:
average error per epoch of module 1 = 0.29996790608531915
TEST:
average weight per epoch of module 1 = 0.21297670070064267
TEST:
average error per epoch of module 2 = 0.3322266611440392
TEST:
average weight per epoch of module 2 = 0.2014497783024335
TEST:
average error per epoch of module 3 = 0.33674762989583235
TEST:
average weight per epoch of module 3 = 0.26253665816305155
TEST:
training all modules and gating network, on all examples (for 1 epoch per
example),
for 50 epochs
m...............................................................................
-EDITED-
..............................
TEST:
average error per epoch of module 1 = 0.4052036345450615
TEST:
average weight per epoch of module 1 = 0.12937120456648465
TEST:
average error per epoch of module 2 = 0.4736577133796439
TEST:
average weight per epoch of module 2 = 0.1417863601427705
TEST:
average error per epoch of module 3 = 0.07887762463444313
TEST:
average weight per epoch of module 3 = 0.7285799989086873
TEST:
training all modules and gating network, on all examples (for 1 epoch per
example),
for 50 epochs
m...............................................................................
-EDITED-
..............................
TEST:
average error per epoch of module 1 = 0.3767399831623017
TEST:
average weight per epoch of module 1 = 0.2147852284321644
TEST:
average error per epoch of module 2 = 0.36932931980672207
TEST:
average weight per epoch of module 2 = 0.23905104781952322
TEST:
average error per epoch of module 3 = 0.42020080128007664
TEST:
average weight per epoch of module 3 = 0.20947016996358023
setting
best modules for each input pattern
input
set 1 best modules are: 2 3 1 3 1
input
set 2 best modules are: 2 3 2 3 1
input
set 3 best modules are: 2 3 3 3 3
input
set 4 best modules are: 2 3 3 3 1
input
set 5 best modules are: 2 3 1 3 3
input
set 6 best modules are: 2 3 3 3 1
input
set 7 best modules are: 3 3 2 3 2
input
set 8 best modules are: 3 3 2 3 2
input
set 9 best modules are: 3 3 2 2 2
input
set 10 best modules are: 3 3 2 3 2
input
set 11 best modules are: 2 3 3 1 1
input
set 12 best modules are: 2 3 3 3 1
input
set 13 best modules are: 2 3 1 1 1
input
set 14 best modules are: 2 3 1 3 1
input
set 15 best modules are: 3 3 1 2 2
input
set 16 best modules are: 3 3 2 3 2
input
set 17 best modules are: 3 3 2 3 2
input
set 18 best modules are: 3 3 2 3 2
input
set 19 best modules are: 2 3 3 1 1
input
set 20 best modules are: 2 3 3 3 1
input
set 21 best modules are: 2 3 3 1 1
input
set 22 best modules are: 2 3 3 3 1
input
set 23 best modules are: 3 3 2 2 2
input
set 24 best modules are: 3 3 2 3 2
input
set 25 best modules are: 3 3 2 2 2
input
set 26 best modules are: 3 3 2 3 2
input
set 27 best modules are: 2 3 1 1 1
input
set 28 best modules are: 2 3 1 3 1
input
set 29 best modules are: 2 3 1 1 1
input
set 30 best modules are: 2 3 1 3 1
input
set 31 best modules are: 3 3 2 2 2
input
set 32 best modules are: 3 3 2 3 2
input
set 33 best modules are: 3 3 2 2 2
input
set 34 best modules are: 3 3 2 3 2
input
set 35 best modules are: 2 1 2 3 2
input
set 36 best modules are: 2 3 1 3 2
input
set 37 best modules are: 2 1 2 3 2
input
set 38 best modules are: 2 3 1 3 2
input
set 39 best modules are: 3 3 1 3 1
input
set 40 best modules are: 3 3 1 3 1
input
set 41 best modules are: 3 3 1 3 1
input
set 42 best modules are: 3 3 1 3 1
input
set 43 best modules are: 2 1 1 3 2
input
set 44 best modules are: 2 3 1 3 2
input
set 45 best modules are: 2 1 1 3 2
input
set 46 best modules are: 2 3 1 3 2
input
set 47 best modules are: 3 3 1 3 1
input
set 48 best modules are: 3 3 1 3 1
input
set 49 best modules are: 3 3 1 3 1
input
set 50 best modules are: 3 3 1 3 1
finding
sub-tasks
module
2 of the first modular network defines sub-task 1
module
3 of the first modular network defines sub-task 2
so
there are 2 sub-tasks
setting
up a breeder for sub-task 1
input
and desired output pattern 1 added as training pattern
module
2 from modular network 1 added to population
module
3 from modular network 2 added to population
module
1 from modular network 3 added to population
module
3 from modular network 4 added to population
module
1 from modular network 5 added to population
input
and desired output pattern 2 added as training pattern
module
2 from modular network 3 added to population
input
and desired output pattern 3 added as training pattern
module
3 from modular network 3 added to population
module
3 from modular network 5 added to population
input
and desired output pattern 4 added as training pattern
input
and desired output pattern 5 added as training pattern
input
and desired output pattern 6 added as training pattern
input
and desired output pattern 11 added as training pattern
module
1 from modular network 4 added to population
input
and desired output pattern 12 added as training pattern
input
and desired output pattern 13 added as training pattern
input
and desired output pattern 14 added as training pattern
input
and desired output pattern 19 added as training pattern
input
and desired output pattern 20 added as training pattern
input
and desired output pattern 21 added as training pattern
input
and desired output pattern 22 added as training pattern
input
and desired output pattern 27 added as training pattern
input
and desired output pattern 28 added as training pattern
input
and desired output pattern 29 added as training pattern
input
and desired output pattern 30 added as training pattern
input
and desired output pattern 35 added as training pattern
module
1 from modular network 2 added to population
module
2 from modular network 5 added to population
input
and desired output pattern 36 added as training pattern
input
and desired output pattern 37 added as training pattern
input
and desired output pattern 38 added as training pattern
input
and desired output pattern 43 added as training pattern
input
and desired output pattern 44 added as training pattern
input
and desired output pattern 45 added as training pattern
input
and desired output pattern 46 added as training pattern
setting
up a breeder for sub-task 2
input
and desired output pattern 7 added as training pattern
module
3 from modular network 1 added to population
module
3 from modular network 2 added to population
module
2 from modular network 3 added to population
module
3 from modular network 4 added to population
module
2 from modular network 5 added to population
input
and desired output pattern 8 added as training pattern
input
and desired output pattern 9 added as training pattern
module
2 from modular network 4 added to population
input
and desired output pattern 10 added as training pattern
input
and desired output pattern 15 added as training pattern
module
1 from modular network 3 added to population
input
and desired output pattern 16 added as training pattern
input
and desired output pattern 17 added as training pattern
input
and desired output pattern 18 added as training pattern
input
and desired output pattern 23 added as training pattern
input
and desired output pattern 24 added as training pattern
input
and desired output pattern 25 added as training pattern
input
and desired output pattern 26 added as training pattern
input
and desired output pattern 31 added as training pattern
input
and desired output pattern 32 added as training pattern
input
and desired output pattern 33 added as training pattern
input
and desired output pattern 34 added as training pattern
input
and desired output pattern 39 added as training pattern
module
1 from modular network 5 added to population
input
and desired output pattern 40 added as training pattern
input
and desired output pattern 41 added as training pattern
input
and desired output pattern 42 added as training pattern
input
and desired output pattern 47 added as training pattern
input
and desired output pattern 48 added as training pattern
input
and desired output pattern 49 added as training pattern
input
and desired output pattern 50 added as training pattern
starting
breeding
TEST:
generation 1 stats
TEST:
population size = 11
TEST:
population member 1 has 0 nodes (0 active)
TEST:
population member 2 has 4 nodes (4 active)
TEST:
population member 3 has 3 nodes (3 active)
TEST:
population member 4 has 0 nodes (0 active)
TEST:
population member 5 has 8 nodes (8 active)
TEST:
population member 6 has 1 nodes (1 active)
TEST:
population member 7 has 1 nodes (1 active)
TEST:
population member 8 has 2 nodes (2 active)
TEST:
population member 9 has 3 nodes (3 active)
TEST:
population member 10 has 1 nodes (1 active)
TEST:
population member 11 has 6 nodes (6 active)
TEST:
starting training population
TEST:
starting training member 1
..................................................
TEST:
starting training member 2
..................................................
TEST:
starting training member 3
..................................................
TEST:
starting training member 4
..................................................
TEST:
starting training member 5
..................................................
TEST:
starting training member 6
..................................................
TEST:
starting training member 7
..................................................
TEST:
starting training member 8
..................................................
TEST:
starting training member 9
..................................................
TEST:
starting training member 10
..................................................
TEST:
starting training member 11
..................................................
TEST:
stopped training population
TEST:
starting evaluating population fitnesses
TEST:
population member 1 error = 2.346548943012448
TEST:
population member 1 fitness = 0.4261577424062683
TEST:
population member 2 error = 6.517898790710071
TEST:
population member 2 fitness = 0.15342367718647226
TEST:
population member 3 error = 8.144048217244848
TEST:
population member 3 fitness = 0.12278905690692271
TEST:
population member 4 error = 2.5337925905126073
TEST:
population member 4 fitness = 0.39466529491969654
TEST:
population member 5 error = 14.66581536234601
TEST:
population member 5 fitness = 0.06818577592129427
TEST:
population member 6 error = 19.540320025086476
TEST:
population member 6 fitness = 0.05117623450978124
TEST:
population member 7 error = 16.506843266315066
TEST:
population member 7 fitness = 0.060580935062288066
TEST:
population member 8 error = 16.968475681267158
TEST:
population member 8 fitness = 0.05893281275135274
TEST:
population member 9 error = 12.462880751323373
TEST:
population member 9 fitness = 0.0802382707460163
TEST:
population member 10 error = 16.319407663943664
TEST:
population member 10 fitness = 0.06127673384919567
TEST:
population member 11 error = 27.703443196022896
TEST:
population member 11 fitness = 0.036096596113495374
TEST:
stopped evaluating population fitnesses
TEST:
member 1 is fittest
TEST:
generating generation 2
TEST:
number of members wanted to cross = 6
TEST:
population members 7 4 selected for crossover
TEST:
child 1 added to new population as member 1
TEST:
child 2 added to new population as member 2
TEST:
population members 2 5 selected for crossover
TEST:
child 1 added to new population as member 3
TEST:
child 2 added to new population as member 4
TEST:
population members 1 9 selected for crossover
TEST:
child 1 added to new population as member 5
TEST:
child 2 added to new population as member 6
TEST:
number of members actually crossed = 6
TEST:
5 members to be kept unchanged
TEST:
member 3 added to new population as member 7
TEST:
member 8 added to new population as member 8
TEST:
member 1 added to new population as member 9
TEST:
member 11 added to new population as member 10
TEST:
member 7 added to new population as member 11
TEST:
6 members (from new population) to be mutated
TEST:
member 6 mutated
TEST:
member 7 mutated
TEST:
layer inserted at 0 with 1 nodes
TEST:
member 5 mutated
TEST:
layer inserted at 1 with 2 nodes
TEST:
member 1 mutated
TEST:
member 3 mutated
TEST:
layer inserted at 1 with 2 nodes
TEST:
member 4 mutated
TEST:
layer inserted at 0 with 34 nodes
TEST:
generation 2 stats
TEST:
population size = 11
TEST:
population member 1 has 0 nodes (0 active)
TEST:
population member 2 has 1 nodes (1 active)
TEST:
population member 3 has 8 nodes (8 active)
TEST:
population member 4 has 6 nodes (6 active)
TEST:
population member 5 has 5 nodes (5 active)
TEST:
population member 6 has 0 nodes (0 active)
TEST:
population member 7 has 3 nodes (3 active)
TEST:
population member 8 has 3 nodes (3 active)
TEST:
population member 9 has 34 nodes (34 active)
TEST:
population member 10 has 6 nodes (6 active)
TEST:
population member 11 has 1 nodes (1 active)
TEST:
starting training population
TEST:
starting training member 1
..................................................
TEST:
starting training member 2
..................................................
TEST:
starting training member 3
..................................................
TEST:
starting training member 4
..................................................
TEST:
starting training member 5
..................................................
TEST:
starting training member 6
..................................................
TEST:
starting training member 7
..................................................
TEST:
starting training member 8
..................................................
TEST:
starting training member 9
..................................................
TEST:
starting training member 10
..................................................
TEST:
starting training member 11
..................................................TEST:
stopped training populat
ion
TEST:
starting evaluating population fitnesses
TEST:
population member 1 error = 1.6395007597359927
TEST:
population member 1 fitness = 0.6099417728608001
TEST:
population member 2 error = 15.858206286967949
TEST:
population member 2 fitness = 0.06305883413950707
TEST:
population member 3 error = 13.5884502361773
TEST:
population member 3 fitness = 0.07359190949808563
TEST:
population member 4 error = 16.47975693760936
TEST:
population member 4 fitness = 0.060680506623119246
TEST:
population member 5 error = 14.685229923473488
TEST:
population member 5 fitness = 0.06809563113489685
TEST:
population member 6 error = 1.5592540839515663
TEST:
population member 6 fitness = 0.6413322949045821
TEST:
population member 7 error = 2.160705008280677
TEST:
population member 7 fitness = 0.462811904525423
TEST:
population member 8 error = 20.419016550626875
TEST:
population member 8 fitness = 0.048973955112901826
TEST:
population member 9 error = 25.816631521900433
TEST:
population member 9 fitness = 0.038734720257818794
TEST:
population member 10 error = 14.157323640600762
TEST:
population member 10 fitness = 0.07063481950304311
TEST:
population member 11 error = 15.858206286967949
TEST:
population member 11 fitness = 0.06305883413950707
TEST:
stopped evaluating population fitnesses
TEST:
member 6 is fittest
TEST:
desired fitness reached
TEST:
generation 1 stats
TEST:
population size = 8
TEST:
population member 1 has 0 nodes (0 active)
TEST:
population member 2 has 4 nodes (4 active)
TEST:
population member 3 has 1 nodes (1 active)
TEST:
population member 4 has 0 nodes (0 active)
TEST:
population member 5 has 6 nodes (6 active)
TEST:
population member 6 has 6 nodes (6 active)
TEST:
population member 7 has 3 nodes (3 active)
TEST:
population member 8 has 8 nodes (8 active)
TEST:
starting training population
TEST:
starting training member 1
..................................................
TEST:
starting training member 2
..................................................
TEST:
starting training member 3
..................................................
TEST:
starting training member 4
..................................................
TEST:
starting training member 5
..................................................
TEST:
starting training member 6
..................................................
TEST:
starting training member 7
..................................................
TEST:
starting training member 8
..................................................
TEST:
stopped training population
TEST:
starting evaluating population fitnesses
TEST:
population member 1 error = 2.068180349714895
TEST:
population member 1 fitness = 0.483516826826951
TEST:
population member 2 error = 11.270487659640729
TEST:
population member 2 fitness = 0.08872730534819441
TEST:
population member 3 error = 12.64265723847817
TEST:
population member 3 fitness = 0.07909729585616548
TEST:
population member 4 error = 2.270577018679533
TEST:
population member 4 fitness = 0.44041668341272816
TEST:
population member 5 error = 10.29069163178764
TEST:
population member 5 fitness = 0.09717519830358436
TEST:
population member 6 error = 26.238531257736096
TEST:
population member 6 fitness = 0.03811188934994838
TEST:
population member 7 error = 12.401509799875937
TEST:
population member 7 fitness = 0.08063534328779903
TEST:
population member 8 error = 6.703212271718719
TEST:
population member 8 fitness = 0.14918220689788744
TEST:
stopped evaluating population fitnesses
TEST:
member 1 is fittest
TEST:
generating generation 2
TEST:
number of members wanted to cross = 4
TEST:
population members 5 1 selected for crossover
TEST:
child 1 added to new population as member 1
TEST:
child 2 added to new population as member 2
TEST:
population members 4 7 selected for crossover
TEST:
child 1 added to new population as member 3
TEST:
child 2 added to new population as member 4
TEST:
number of members actually crossed = 4
TEST:
4 members to be kept unchanged
TEST:
member 1 added to new population as member 5
TEST:
member 8 added to new population as member 6
TEST:
member 4 added to new population as member 7
TEST:
member 6 added to new population as member 8
TEST:
4 members (from new population) to be mutated
TEST:
member 4 mutated
TEST:
member 5 mutated
TEST:
node 3 removed from layer 2
TEST:
member 2 mutated
TEST:
connection 3 removed from output 2
TEST:
member 1 mutated
TEST:
generation 2 stats
TEST:
population size = 8
TEST:
population member 1 has 0 nodes (0 active)
TEST:
population member 2 has 6 nodes (6 active)
TEST:
population member 3 has 3 nodes (3 active)
TEST:
population member 4 has 0 nodes (0 active)
TEST:
population member 5 has 0 nodes (0 active)
TEST:
population member 6 has 7 nodes (7 active)
TEST:
population member 7 has 0 nodes (0 active)
TEST:
population member 8 has 6 nodes (6 active)
TEST:
starting training population
TEST:
starting training member 1
..................................................
TEST:
starting training member 2
..................................................
TEST:
starting training member 3
..................................................
TEST:
starting training member 4
..................................................
TEST:
starting training member 5
..................................................
TEST:
starting training member 6
..................................................
TEST:
starting training member 7
..................................................
TEST:
starting training member 8
..................................................TEST:
stopped training populat
ion
TEST:
starting evaluating population fitnesses
TEST:
population member 1 error = 1.370242384324583
TEST:
population member 1 fitness = 0.7297978893660613
TEST:
population member 2 error = 13.266858121273074
TEST:
population member 2 fitness = 0.07537579665501398
TEST:
population member 3 error = 12.222886296452046
TEST:
population member 3 fitness = 0.0818137366041171
TEST:
population member 4 error = 1.455161876371046
TEST:
population member 4 fitness = 0.6872087677927963
TEST:
population member 5 error = 1.370242384324583
TEST:
population member 5 fitness = 0.7297978893660613
TEST:
population member 6 error = 12.152474417979121
TEST:
population member 6 fitness = 0.0822877683676123
TEST:
population member 7 error = 1.455161876371046
TEST:
population member 7 fitness = 0.6872087677927963
TEST:
population member 8 error = 26.22730347208864
TEST:
population member 8 fitness = 0.038128204871088256
TEST:
stopped evaluating population fitnesses
TEST:
member 1 is fittest
TEST:
desired fitness reached
setting
modular networks from breeders
taking
members from breeder 1
population
member 6 added as module in modular network 1
population
member 1 added as module in modular network 2
population
member 7 added as module in modular network 3
population
member 3 added as module in modular network 4
population
member 10 added as module in modular network 5
taking
members from breeder 2
population
member 1 added as module in modular network 1
population
member 5 added as module in modular network 2
population
member 4 added as module in modular network 3
population
member 7 added as module in modular network 4
population
member 6 added as module in modular network 5
finding
best modular network
the
error of modular network 1 = 0.15119914030554285
the
error of modular network 2 = 0.10739070076224232
the
error of modular network 3 = 0.1353945574954972
the
error of modular network 4 = 0.1708024881067879
the
error of modular network 5 = 0.25144689163660056
best
modular network = 2
training
all modular networks
TEST:
training all modules and gating network, on all examples (for 1 epoch per
example),
for 50 epochs
m...............................................................................
-EDITED-
..........
TEST:
average error per epoch of module 1 = 0.06346829607325283
TEST:
average weight per epoch of module 1 = 0.3882876310903775
TEST:
average error per epoch of module 2 = 0.07568032669363639
TEST:
average weight per epoch of module 2 = 0.3663402007494739
TEST:
training all modules and gating network, on all examples (for 1 epoch per
example),
for 50 epochs
m...............................................................................
-EDITED-
..........
TEST:
average error per epoch of module 1 = 0.008956695295416189
TEST:
average weight per epoch of module 1 = 0.416072675051528
TEST:
average error per epoch of module 2 = 0.08575067360454604
TEST:
average weight per epoch of module 2 = 0.2416748617554456
TEST:
training all modules and gating network, on all examples (for 1 epoch per
example),
for 50 epochs
m...............................................................................
-EDITED-
..........
TEST:
average error per epoch of module 1 = 0.06621125679930567
TEST:
average weight per epoch of module 1 = 0.4264438882887907
TEST:
average error per epoch of module 2 = 0.022081959444451935
TEST:
average weight per epoch of module 2 = 0.2990247137944984
TEST:
training all modules and gating network, on all examples (for 1 epoch per
example),
for 50 epochs
m...............................................................................
-EDITED-
..........
TEST:
average error per epoch of module 1 = 0.27410735096473654
TEST:
average weight per epoch of module 1 = 0.10882186324086639
TEST:
average error per epoch of module 2 = 0.007186064153273938
TEST:
average weight per epoch of module 2 = 0.7296107363617412
TEST:
training all modules and gating network, on all examples (for 1 epoch per
example),
for 50 epochs
m...............................................................................
-EDITED-
..........
TEST:
average error per epoch of module 1 = 0.20434542089946361
TEST:
average weight per epoch of module 1 = 0.28501589628080753
TEST:
average error per epoch of module 2 = 0.2261209784751715
TEST:
average weight per epoch of module 2 = 0.26803929014780775
setting
best modules for each input pattern
input
set 1 best modules are: 1 1 2 2 1
input
set 2 best modules are: 1 1 2 2 2
input
set 3 best modules are: 1 1 1 2 1
input
set 4 best modules are: 1 1 1 2 1
input
set 5 best modules are: 1 1 1 2 1
input
set 6 best modules are: 1 1 1 2 1
input
set 7 best modules are: 2 1 2 2 2
input
set 8 best modules are: 2 2 2 2 2
input
set 9 best modules are: 2 2 2 2 2
input
set 10 best modules are: 2 2 2 2 2
input
set 11 best modules are: 1 1 1 2 2
input
set 12 best modules are: 1 1 1 2 1
input
set 13 best modules are: 1 1 1 2 1
input
set 14 best modules are: 1 1 1 2 1
input
set 15 best modules are: 2 2 2 2 2
input
set 16 best modules are: 2 2 2 2 2
input
set 17 best modules are: 1 2 2 2 2
input
set 18 best modules are: 2 2 2 2 2
input
set 19 best modules are: 2 1 1 2 2
input
set 20 best modules are: 1 1 1 2 1
input
set 21 best modules are: 1 1 1 2 2
input
set 22 best modules are: 1 1 1 2 1
input
set 23 best modules are: 1 2 2 2 2
input
set 24 best modules are: 2 2 2 2 2
input
set 25 best modules are: 2 2 2 2 2
input
set 26 best modules are: 2 2 2 2 2
input
set 27 best modules are: 2 1 1 2 2
input
set 28 best modules are: 1 1 1 2 1
input
set 29 best modules are: 2 1 1 2 2
input
set 30 best modules are: 1 1 1 2 1
input
set 31 best modules are: 1 2 2 2 2
input
set 32 best modules are: 2 2 2 2 2
input
set 33 best modules are: 1 2 2 2 2
input
set 34 best modules are: 2 2 2 2 2
input
set 35 best modules are: 1 1 1 2 1
input
set 36 best modules are: 1 1 1 2 1
input
set 37 best modules are: 1 1 1 2 1
input
set 38 best modules are: 1 1 1 2 1
input
set 39 best modules are: 2 2 2 2 2
input
set 40 best modules are: 2 2 2 2 2
input
set 41 best modules are: 2 2 2 2 1
input
set 42 best modules are: 2 2 2 2 2
input
set 43 best modules are: 1 1 1 2 1
input
set 44 best modules are: 1 1 1 2 1
input
set 45 best modules are: 1 1 1 2 1
input
set 46 best modules are: 1 1 1 2 1
input
set 47 best modules are: 2 2 2 2 1
input
set 48 best modules are: 2 2 2 2 2
input
set 49 best modules are: 2 2 2 2 1
input
set 50 best modules are: 2 2 2 2 2
finding
sub-tasks
module
1 of the first modular network defines sub-task 1
module
2 of the first modular network defines sub-task 2
so
there are 2 sub-tasks
setting
up a breeder for sub-task 1
input
and desired output pattern 1 added as training pattern
module
1 from modular network 1 added to population
module
1 from modular network 2 added to population
module
2 from modular network 3 added to population
module
2 from modular network 4 added to population
module
1 from modular network 5 added to population
input
and desired output pattern 2 added as training pattern
module
2 from modular network 5 added to population
input
and desired output pattern 3 added as training pattern
module
1 from modular network 3 added to population
input
and desired output pattern 4 added as training pattern
input
and desired output pattern 5 added as training pattern
input
and desired output pattern 6 added as training pattern
input
and desired output pattern 11 added as training pattern
input
and desired output pattern 12 added as training pattern
input
and desired output pattern 13 added as training pattern
input
and desired output pattern 14 added as training pattern
input
and desired output pattern 17 added as training pattern
module
2 from modular network 2 added to population
input
and desired output pattern 20 added as training pattern
input
and desired output pattern 21 added as training pattern
input
and desired output pattern 22 added as training pattern
input
and desired output pattern 23 added as training pattern
input
and desired output pattern 28 added as training pattern
input
and desired output pattern 30 added as training pattern
input
and desired output pattern 31 added as training pattern
input
and desired output pattern 33 added as training pattern
input
and desired output pattern 35 added as training pattern
input
and desired output pattern 36 added as training pattern
input
and desired output pattern 37 added as training pattern
input
and desired output pattern 38 added as training pattern
input
and desired output pattern 43 added as training pattern
input
and desired output pattern 44 added as training pattern
input
and desired output pattern 45 added as training pattern
input
and desired output pattern 46 added as training pattern
setting
up a breeder for sub-task 2
input
and desired output pattern 7 added as training pattern
module
2 from modular network 1 added to population
module
1 from modular network 2 added to population
module
2 from modular network 3 added to population
module
2 from modular network 4 added to population
module
2 from modular network 5 added to population
input
and desired output pattern 8 added as training pattern
module
2 from modular network 2 added to population
input
and desired output pattern 9 added as training pattern
input
and desired output pattern 10 added as training pattern
input
and desired output pattern 15 added as training pattern
input
and desired output pattern 16 added as training pattern
input
and desired output pattern 18 added as training pattern
input
and desired output pattern 19 added as training pattern
module
1 from modular network 3 added to population
input
and desired output pattern 24 added as training pattern
input
and desired output pattern 25 added as training pattern
input
and desired output pattern 26 added as training pattern
input
and desired output pattern 27 added as training pattern
input
and desired output pattern 29 added as training pattern
input
and desired output pattern 32 added as training pattern
input
and desired output pattern 34 added as training pattern
input
and desired output pattern 39 added as training pattern
input
and desired output pattern 40 added as training pattern
input
and desired output pattern 41 added as training pattern
module
1 from modular network 5 added to population
input
and desired output pattern 42 added as training pattern
input
and desired output pattern 47 added as training pattern
input
and desired output pattern 48 added as training pattern
input
and desired output pattern 49 added as training pattern
input
and desired output pattern 50 added as training pattern
starting
breeding
TEST:
generation 1 stats
TEST:
population size = 8
TEST:
population member 1 has 0 nodes (0 active)
TEST:
population member 2 has 0 nodes (0 active)
TEST:
population member 3 has 0 nodes (0 active)
TEST:
population member 4 has 0 nodes (0 active)
TEST:
population member 5 has 6 nodes (6 active)
TEST:
population member 6 has 7 nodes (7 active)
TEST:
population member 7 has 3 nodes (3 active)
TEST:
population member 8 has 0 nodes (0 active)
TEST:
starting training population
TEST:
starting training member 1
..................................................
TEST:
starting training member 2
..................................................
TEST:
starting training member 3
..................................................
TEST:
starting training member 4
..................................................
TEST:
starting training member 5
..................................................
TEST:
starting training member 6
..................................................
TEST:
starting training member 7
..................................................
TEST:
starting training member 8
..................................................
TEST:
stopped training population
TEST:
starting evaluating population fitnesses
TEST:
population member 1 error = 1.3523475975575454
TEST:
population member 1 fitness = 0.7394548574686604
TEST:
population member 2 error = 1.1287202030578243
TEST:
population member 2 fitness = 0.885959157363262
TEST:
population member 3 error = 1.5681119470500533
TEST:
population member 3 fitness = 0.6377095728919158
TEST:
population member 4 error = 1.2712865802366433
TEST:
population member 4 fitness = 0.7866047007385661
TEST:
population member 5 error = 14.034910805868503
TEST:
population member 5 fitness = 0.0712508981234041
TEST:
population member 6 error = 15.452216914005213
TEST:
population member 6 fitness = 0.06471563307486602
TEST:
population member 7 error = 0.9571828247056348
TEST:
population member 7 fitness = 1.044732494346138
TEST:
population member 8 error = 2.301361722269734
TEST:
population member 8 fitness = 0.43452534659077546
TEST:
stopped evaluating population fitnesses
TEST:
member 7 is fittest
TEST:
desired fitness reached
TEST:
generation 1 stats
TEST:
population size = 8
TEST:
population member 1 has 0 nodes (0 active)
TEST:
population member 2 has 0 nodes (0 active)
TEST:
population member 3 has 0 nodes (0 active)
TEST:
population member 4 has 0 nodes (0 active)
TEST:
population member 5 has 7 nodes (7 active)
TEST:
population member 6 has 0 nodes (0 active)
TEST:
population member 7 has 3 nodes (3 active)
TEST:
population member 8 has 6 nodes (6 active)
TEST:
starting training population
TEST:
starting training member 1
..................................................
TEST:
starting training member 2
..................................................
TEST:
starting training member 3
..................................................
TEST:
starting training member 4
..................................................
TEST:
starting training member 5
..................................................
TEST:
starting training member 6
..................................................
TEST:
starting training member 7
..................................................
TEST:
starting training member 8
..................................................
TEST:
stopped training population
TEST:
starting evaluating population fitnesses
TEST:
population member 1 error = 1.131799873450384
TEST:
population member 1 fitness = 0.8835484288855932
TEST:
population member 2 error = 1.162418027681366
TEST:
population member 2 fitness = 0.8602757150924994
TEST:
population member 3 error = 0.9463779972273676
TEST:
population member 3 fitness = 1.056660238223765
TEST:
population member 4 error = 0.9038187612766001
TEST:
population member 4 fitness = 1.1064165105264563
TEST:
population member 5 error = 10.251427033112359
TEST:
population member 5 fitness = 0.09754739479391265
TEST:
population member 6 error = 1.2020396423173496
TEST:
population member 6 fitness = 0.8319193184612049
TEST:
population member 7 error = 1.6448340802542987
TEST:
population member 7 fitness = 0.6079640566818725
TEST:
population member 8 error = 17.529398885390552
TEST:
population member 8 fitness = 0.05704702177970435
TEST:
stopped evaluating population fitnesses
TEST: member 4