def training_ANN(self):
# Normalize because collective signals have E so high
init.INPUT_DATASETs = np.divide(init.INPUT_DATASETs, 500)
self.NN = ann.Neural_Network(Lambda = 0.0001)
self.T = ann.trainer(self.NN)
self.T.train(init.INPUT_DATASETs, init.OUTPUT_DATASETs)
''' Draw training data, relation between T and error E '''
plt.figure(1)
plt.plot(self.T.E, label = 'Train line', linewidth = 2.0)
plt.legend()
plt.grid(1)
plt.xlabel('Epochs')
plt.ylabel('Cost')
plt.show()
temp_accuracy = []
count1 = 0
count2 = 0
count3 = 0
count4 = 0
count = 0
temp_accuracy_test = []
count1_test = 0
count2_test = 0
count3_test = 0
count4_test = 0
count_test = 0
''' Training ANN '''
NN = ANN.Neural_Network(Lambda = 0.0001)
T = ANN.trainer(NN)
T.train(init.INPUT_DATASETs, init.OUTPUT_DATASETs)
endTime = time.clock() # Get end time
# Calculate processing time
processing_time = startTime - endTime
test = NN.foward(init.INPUT_DATASETs)
test_test = NN.foward(INPUT_DATASETs_test)
# Accuracy of UP state
for t in range(0, test.shape[0]/4):
if test[t][0] == np.max(test[t]):
count1 += 1
for t in range(0, test_test.shape[0]/4):
if test_test[t][0] == np.max(test_test[t]):
temp_accuracy = []
count1 = 0
count2 = 0
count3 = 0
count4 = 0
count = 0
count1_test = 0
count2_test = 0
count3_test = 0
count4_test = 0
count_test = 0
startTime1 = time.clock()
''' Training ANN '''
NN_UP = ANN.Neural_Network(Lambda=0.0001)
T_UP = ANN.trainer(NN_UP)
T_UP.train(init.UP_INPUT, init.UP_OUTPUT)
# RIGHT Neural Nets
NN_RIGHT = ANN.Neural_Network(Lambda=0.0001)
T_RIGHT = ANN.trainer(NN_RIGHT)
T_RIGHT.train(init.RIGHT_INPUT, init.RIGHT_OUTPUT)
# DOWN Neural Nets
NN_DOWN = ANN.Neural_Network(Lambda=0.0001)
T_DOWN = ANN.trainer(NN_DOWN)
T_DOWN.train(init.DOWN_INPUT, init.DOWN_OUTPUT)
# LEFT Neural Nets
NN_LEFT = ANN.Neural_Network(Lambda=0.0001)
