def step_statictics(simu, network, plot, inputs, outputs):
cell = [0 for _ in range(10)]
for k in range(10):
if index_max_nth(network['FoN'].stateOutputNeurons,
k) == index_max(outputs):
cell[k] = 1
break
#rms
simu.rms('FoN', inputs, outputs)
simu.rms('SoN', network['FoN'].stateHiddenNeurons, cell)
#err
simu.perf('FoN', outputs)
simu.perf('SoN', cell)
#wager ratio
if (index_max(network['SoN'].stateOutputNeurons) == 0):
plot['high_wager'] += 1
#feedback
if (index_max_nth(
network['FoN'].stateOutputNeurons,
index_max(
network['SoN'].stateOutputNeurons)) == index_max(outputs)):
plot['feedback'] += 1
def step_statictics(simu, network, plot, inputs, outputs):
cell = [0., 0.]
if index_max(network['FoN'].stateOutputNeurons) == index_max(outputs):
cell = [0., 1]
else:
cell = [1, 0.]
#rms
simu.rms('FoN', inputs, outputs)
simu.rms('SoN', network['FoN'].stateHiddenNeurons, cell)
#err
simu.perf('FoN', outputs)
simu.perf('SoN', cell)
#wager ratio
if(index_max(network['SoN'].stateOutputNeurons) == 1):
plot['high_wager'] += 1
#feedback
if(index_max(network['SoN'].stateOutputNeurons) == 1):
if(index_max(network['FoN'].stateOutputNeurons) == index_max(outputs)):
plot['feedback'] += 1
if(index_max(network['SoN'].stateOutputNeurons) == 0):
if(index_max_nth(network['FoN'].stateOutputNeurons,1) == index_max(outputs)):
plot['feedback'] += 1
def step_statictics(simu, network, plot, inputs, outputs):
cell = [0., 0.]
if index_max(network['FoN'].stateOutputNeurons) == index_max(outputs):
cell = [0., 1]
else:
cell = [1, 0.]
#rms
simu.rms('FoN', inputs, outputs)
simu.rms('SoN', network['FoN'].stateHiddenNeurons, cell)
#err
simu.perf('FoN', outputs)
simu.perf('SoN', cell)
#wager ratio
if (index_max(network['SoN'].stateOutputNeurons) == 1):
plot['high_wager'] += 1
#feedback
if (index_max(network['SoN'].stateOutputNeurons) == 1):
if (index_max(
network['FoN'].stateOutputNeurons) == index_max(outputs)):
plot['feedback'] += 1
if (index_max(network['SoN'].stateOutputNeurons) == 0):
if (index_max_nth(network['FoN'].stateOutputNeurons,
1) == index_max(outputs)):
plot['feedback'] += 1
def step_learn(network, inputs, outputs):
cell = [0 for _ in range(10)]
for k in range(10):
if index_max_nth(network['FoN'].stateOutputNeurons, k) == index_max(outputs):
cell[k] = 1
break
#Learning
network['SoN'].train(network['FoN'].stateHiddenNeurons, cell)
network['FoN'].train(inputs, outputs)
def step_learn(network, inputs, outputs):
cell = [0 for _ in range(10)]
for k in range(10):
if index_max_nth(network['FoN'].stateOutputNeurons,
k) == index_max(outputs):
cell[k] = 1
break
#Learning
network['SoN'].train(network['FoN'].stateHiddenNeurons, cell)
network['FoN'].train(inputs, outputs)
def step_statictics(simu, network, plot, inputs, outputs):
cell = [0 for _ in range(10)]
for k in range(10):
if index_max_nth(network['FoN'].stateOutputNeurons, k) == index_max(outputs):
cell[k] = 1
break
#rms
simu.rms('FoN', inputs, outputs)
simu.rms('SoN', network['FoN'].stateHiddenNeurons, cell)
#err
simu.perf('FoN', outputs)
simu.perf('SoN', cell)
#wager ratio
if(index_max(network['SoN'].stateOutputNeurons) == 0):
plot['high_wager'] += 1
#feedback
if(index_max_nth(network['FoN'].stateOutputNeurons, index_max(network['SoN'].stateOutputNeurons)) == index_max(outputs)):
plot['feedback'] += 1