Hof L reinventing motorbabbling :)
http://twitter.com/chris_brockett/status/169896605556813824
http://www.cs.uvm.edu/~jbongard/papers/2006_Science_Bongard_Zykov_Lipson.pdf
http://www.sciencemag.org/content/314/5802/1118/rel-suppl/104c9b29bce86db6/suppl/DC1
Correction, actually I think instead I'll try again to install Theano and get the DBN working with the motor babbling. This may identify interesting representations which fitness can be a function of.
OK, so here is an example time-series from one of the motor actions.
Here are some papers that might tell me how to analyse the above time series to discover "interesting actions".
[Sensor1, Sensor2, Motor, SensorP]
[ 1.60739040e+00 9.66061115e+00 6.17505843e-03 1.01002534e+00]
[ 1.61297333e+00 9.59638214e+00 6.17507054e-03 1.00988591e+00]
[ 1.61431706e+00 9.71891022e+00 6.17509242e-03 1.01107304e+00]
[ 1.60549104e+00 9.69761848e+00 6.17510499e-03 1.01021157e+00]
[ 1.61960542e+00 9.71553326e+00 6.17511431e-03 1.01144421e+00]
[ 1.59876752e+00 9.59617138e+00 6.17511664e-03 1.00878826e+00]
[ 1.61962616e+00 9.71207047e+00 6.17511757e-03 1.01141508e+00]
[ 1.62076795e+00 9.64346123e+00 6.17511803e-03 1.01089748e+00]
[ 1.61177838e+00 9.71861362e+00 6.17511757e-03 1.01087776e+00]
[ 1.60186958e+00 9.60560226e+00 6.17511664e-03 1.00911151e+00]
and
[ 0.04912947 0. 1.04316938 0.12033726]
[ 0.04912947 0. 1.04316938 0.12033726]
[ 0.04912947 0. 1.04316938 0.12033726]
[ 0.04912948 0. 1.04316938 0.12033726]
[ 0.04912948 0. 1.04316938 0.12033726]
[ 0.04912948 0. 1.04316938 0.12033726]
[ 0.04912948 0. 1.04316938 0.12033726]
[ 0.04912948 0. 1.04316938 0.12033726]
[ 0.04912948 0. 1.04316938 0.12033726]
[ 0.04912949 0. 1.04316938 0.12033726]
often there is no change, because stationary joints are chosen, and so the output signal of joint angle is just fixed as well. Probably using a randomly initialised FNN in this way is not a very interesting way of specifying a reactive behaviour. Other closed-loop action representations would be e.g.
1. Readouts from a LSM (Liquid state machine) into which the sensory inputs were being injected.
2. An option (from Sutton's Hierarchical reinforcement learning framework) consisting of a starting sensory condition, and then a behavioural sequence of actions (perhaps including other options) and a termination condition (e.g. sensory or motor state, or time).
Anyway, once a good motor representation has been decided on, we've still got to look at the time series and decide what is a fit time series. It is one in which the motor actions carry lots of information about the sensory states I think. So we need to use an information theory measure of this. The simplest thing to do is to use binning, i.e. break sensory activity into bins of 0.1 units and motor activity into the same bins. Then check the mutual information between preceding timestep motor activities and subsequent time step sensory activities, or even points at the same time step.
This was achieved fairly easily using the pyentropy library and the code below...
#Fitness function (4) ******************************************
#Mutual Information between motor command and predicted sensory state.
sp = smMatrix[1:,3]
mot = smMatrix[1:,2]
spQ = ent.quantise(sp,3)
motQ = ent.quantise(mot,3)
s = ent.DiscreteSystem(spQ[0],(1,3), motQ[0],(1,3))
print(str(spQ[0]) + "\n" + str(motQ[0]))
s.calculate_entropies(method='plugin', calc=['HX', 'HXY'])
mutInf = s.I()
fit = mutInf
print(fit)
return fit
Overnight the following fitness improvement was obtained. Here are some examples of the discretized time series of sm measurements (only 10 points) and their fitness (MI) which is the maximum possible because each category has a one to one correspondence with its counterpart in the other time series.
0 2 0 0 2 2 1 1 1 0]
[0 2 0 0 2 2 1 1 1 0]
1.57095059445
[2 0 0 0 0 2 2 1 1 1]
[2 0 0 0 0 2 2 1 1 1]
1.57095059445
[2 2 0 0 0 1 2 1 1 0]
[2 2 0 0 0 1 2 1 1 0]
1.57095059445
[2 2 0 0 0 2 1 1 1 0]
[2 2 0 0 0 2 1 1 1 0]
1.57095059445
[1 0 0 2 2 2 1 1 0 0]
[1 0 0 2 2 2 1 1 0 0]
1.57095059445
[2 2 0 0 0 2 1 1 1 0]
[2 2 0 0 0 2 1 1 1 0]
1.57095059445
[2 2 2 1 1 0 0 0 0 1]
[2 2 2 1 1 0 0 0 0 1]
1.57095059445
[1 2 2 2 0 1 1 0 0 0]
[1 2 2 2 0 1 1 0 0 0]
1.57095059445
This max MI was achieved very rapidly after only a few generations. After 493 generations here are some examples of the genotypes in individuals from 0 to 4 and from 35 to 49.
Gen = 493,
Num = 0,
Fit = 1.57095059445,
Genotype
Sens1, Sens2 in = [[27, 37],
Mot = 2,
Predicted Sens = 37,
Weights = [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 1, 1.57095059445,
S1,S2 = [[27, 37],
Mot = 2,
PredS = 37,
[-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 2, 1.57095059445, [[27, 37],
Mot = 13,
PredS = 37,
[-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 3, 1.57095059445, [[27, 37],
Mot = 21,
PredS = 37, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 4, 1.57095059445, [[27, 37],
Mot = 13,
PredS = 37,
[-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 28, 1.57095059445, [[27, 37],
Mot = 25,
PredS = 14, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
Gen = 493,
Indiv number = 35,
1.24643934467, [[27, 37], 25, 42, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 36, 0.970950594455, [[11, 43], 6, 37, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 37, 0.646439344671, [[27, 37], 2, 3, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 38, 0.619973094022, [[27, 37], 19, 15, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 39, 0.419973094022, [[12, 0], 12, 37, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 40, 0.419973094022, [[30, 36], 14, 37, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 41, 0.419973094022, [[27, 37], 14, 40, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 42, 0.193506843373, [[27, 37], 14, 3, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 43, 0.144484343806, [[23, 27], 13, 37, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 44, 0.0, [[27, 37], 25, 23, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 45, 0.0, [[27, 37], 13, 25, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 46, 0.0, [[27, 37], 13, 33, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 47, -0.0, [[33, 33], 22, 37, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 48, 0.0, [[27, 37], 19, 6, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
493, 49, 0.0, [[27, 37], 13, 35, [-0.7813761451611838, -0.6925101339769224, -0.8882461218366977, 0.5827564911597551, 0.3156858421564597, 0.868992037003147, 0.39791058610088115, 0.7477408128370153, -0.46195335809234184, 0.11317841235128778, 0.3632540465168763, 0.3605233296179058, -0.009367787989753351, -0.9127791268799212, -0.5161354969487246, -0.2819743414778495, -0.7099409667199355, 0.4312249539044628, -0.898855785363035, 0.6726487861687325, 0.18854325800078375, 0.49381892334496813, 0.7564868523037998, -0.8266019644001747, -0.363077383522052, 0.7033604367285744, 0.515903605148013, -0.6067226594500343, -0.7330732058230809, 0.7384439022826668, -0.3945938050605289, -0.10465336072532994, 0.4356088515521128, -0.7276146053576384, 0.15492702273197767, 0.23872287208532916, -0.7942722066925845, -0.03916449666512456, -0.9122418072544121, -0.6426927606848216, -0.6696745736460405]]
The robot moves its arms and legs and feet by large amounts, bilaterally. There are about 5-6 stable movement forms at the end of this 490 generations it seems, all with the maximum possible fitness given this level of discretization.
What is sensor 37?
"Device/SubDeviceList/InertialSensor/GyroscopeX/Sensor/Value"
This is interesting because many of the motors would be expected to have an effect on the gyroscope. It is a convenient and easy sm contingency to discover.
A problem is the current upper ceiling on fitness which is reached quickly with this level of discretization. Perhaps the level of discretization should slowly be increased over generations?
http://www.cs.cas.cz/mp/papers07/causality-IT-final.pdf
No comments:
Post a Comment