noOfSamples = deeplearn.readCsvFile("xor.data", 4, 1, [2], 0) print(str(noOfSamples) + " samples loaded") # The error threshold (percent) for each layer of the network. # After going below the threshold the pre-training will move # on to the next layer deeplearn.setErrorThresholds([0.8, 0.8]) # The learning rate in the range 0.0-1.0 deeplearn.setLearningRate(0.2) # The percentage of dropouts in the range 0-100 deeplearn.setDropoutsPercent(0.001) # Title of the training error image deeplearn.setPlotTitle("Xor Training") # The number of time steps after which the training error image is redrawn deeplearn.setHistoryPlotInterval(50000) print("Training started") timeStep = 0 while (deeplearn.training() != 0): timeStep = timeStep + 1 print("Training Completed") print("Test data set performance is " + str(deeplearn.getPerformance()) + "%") deeplearn.export("result.py") print("Exported trained network")
noOfSamples = deeplearn.readCsvFile("xor.data", 4, 1, [2], 0) print str(noOfSamples) + " samples loaded" # The error threshold (percent) for each layer of the network. # After going below the threshold the pre-training will move # on to the next layer deeplearn.setErrorThresholds([0.8, 0.8]) # The learning rate in the range 0.0-1.0 deeplearn.setLearningRate(0.2) # The percentage of dropouts in the range 0-100 deeplearn.setDropoutsPercent(0.001) # Title of the training error image deeplearn.setPlotTitle("Xor Training") # The number of time steps after which the training error image is redrawn deeplearn.setHistoryPlotInterval(50000) print "Training started" timeStep = 0 while (deeplearn.training() != 0): timeStep = timeStep + 1 print "Training Completed" print "Test data set performance is " + str(deeplearn.getPerformance()) + "%"; deeplearn.export("result.py") print "Exported trained network"
noOfSamples = deeplearn.readCsvFile("wdbc.data", 16, 3, [1], 0) print(str(noOfSamples) + " samples loaded") # The error threshold (percent) for each layer of the network. # After going below the threshold the pre-training will move # on to the next layer deeplearn.setErrorThresholds([1.0, 1.0, 1.5, 6.0]) # The learning rate in the range 0.0-1.0 deeplearn.setLearningRate(0.2) # The percentage of dropouts in the range 0-100 deeplearn.setDropoutsPercent(2.0) # Title of the training error image deeplearn.setPlotTitle("Cancer Classification Training") # The number of time steps after which the training error image is redrawn deeplearn.setHistoryPlotInterval(500000) print("Training started") timeStep = 0 while (deeplearn.training() != 0): timeStep = timeStep + 1 print("Training Completed") print("Test data set performance is " + str(deeplearn.getPerformance()) + "%") deeplearn.export("result.py") print("Exported trained network")
print(str(noOfSamples) + " samples loaded") # The error threshold (percent) for each layer of the network. # Three hidden layers, plus the final training. # After going below the threshold the pre-training will move # on to the next layer deeplearn.setErrorThresholds([1.6, 2.05, 4.0, 9.5]) # The learning rate in the range 0.0-1.0 deeplearn.setLearningRate(0.5) # The percentage of dropouts in the range 0-100 deeplearn.setDropoutsPercent(0.001) # Title of the training error image deeplearn.setPlotTitle("White Wine Quality Training") # The number of time steps after which the training error image is redrawn deeplearn.setHistoryPlotInterval(500000) print("Training started") timeStep = 0 while (deeplearn.training() != 0): timeStep = timeStep + 1 print("Training Completed") print("Test data set performance is " + str(deeplearn.getPerformance()) + "%") deeplearn.export("result.py") print("Exported trained network")
noOfSamples = deeplearn.readCsvFile("iris.data", 16, 3, [4], no_of_classes) print str(noOfSamples) + " samples loaded" # The error threshold (percent) for each layer of the network. # After going below the threshold the pre-training will move # on to the next layer deeplearn.setErrorThresholds([0.5, 0.5, 0.5, 2.5]) # The learning rate in the range 0.0-1.0 deeplearn.setLearningRate(0.1) # The percentage of dropouts in the range 0-100 deeplearn.setDropoutsPercent(0.01) # Title of the training error image deeplearn.setPlotTitle("Iris Species Classification Training") # The number of time steps after which the training error image is redrawn deeplearn.setHistoryPlotInterval(500000) print "Training started" timeStep = 0 while (deeplearn.training() != 0): timeStep = timeStep + 1 print "Training Completed" print "Test data set performance is " + str(deeplearn.getPerformance()) + "%"; deeplearn.export("result.py") print "Exported trained network"
print str(noOfSamples) + " samples loaded" # The error threshold (percent) for each layer of the network. # Three hidden layers, plus the final training. # After going below the threshold the pre-training will move # on to the next layer deeplearn.setErrorThresholds([0.5, 0.5, 0.5, 2.0]) # The learning rate in the range 0.0-1.0 deeplearn.setLearningRate(0.2) # The percentage of dropouts in the range 0-100 deeplearn.setDropoutsPercent(0.001) # Title of the training error image deeplearn.setPlotTitle("Concrete Slump Training") # The number of time steps after which the training error image is redrawn deeplearn.setHistoryPlotInterval(900000) print "Training started" timeStep = 0 while (deeplearn.training() != 0): timeStep = timeStep + 1 print "Training Completed" print "Test data set performance is " + str(deeplearn.getPerformance()) + "%" deeplearn.export("result.py") print "Exported trained network"
noOfSamples = deeplearn.readCsvFile("data.csv", samplesPerAxis*(axes-1), 3, [1], 0) print str(noOfSamples) + " samples loaded" # The error threshold (percent) for each layer of the network. # After going below the threshold the pre-training will move # on to the next layer deeplearn.setErrorThresholds([5.0, 7.0, 7.0, 20.0]) # The learning rate in the range 0.0-1.0 deeplearn.setLearningRate(0.5) # The percentage of dropouts in the range 0-100 deeplearn.setDropoutsPercent(0.001) # Title of the training error image deeplearn.setPlotTitle("Catmuzzle Training") # The number of time steps after which the training error image is redrawn deeplearn.setHistoryPlotInterval(50000) print "Training started" timeStep = 0 while (deeplearn.training() != 0): timeStep = timeStep + 1 print "Training Completed" print "Test data set performance is " + str(deeplearn.getPerformance()) + "%"; deeplearn.export("result.py") print "Exported trained network"
noOfSamples = deeplearn.readCsvFile("wdbc.data", 16, 3, [1], 0) print str(noOfSamples) + " samples loaded" # The error threshold (percent) for each layer of the network. # After going below the threshold the pre-training will move # on to the next layer deeplearn.setErrorThresholds([3.0, 3.0, 3.5, 3.0]) # The learning rate in the range 0.0-1.0 deeplearn.setLearningRate(0.2) # The percentage of dropouts in the range 0-100 deeplearn.setDropoutsPercent(2.0) # Title of the training error image deeplearn.setPlotTitle("Cancer Classification Training") # The number of time steps after which the training error image is redrawn deeplearn.setHistoryPlotInterval(500000) print "Training started" timeStep = 0 while (deeplearn.training() != 0): timeStep = timeStep + 1 print "Training Completed" print "Test data set performance is " + str(deeplearn.getPerformance()) + "%"; deeplearn.export("result.py") print "Exported trained network"
print str(noOfSamples) + " samples loaded" # The error threshold (percent) for each layer of the network. # Three hidden layers, plus the final training. # After going below the threshold the pre-training will move # on to the next layer deeplearn.setErrorThresholds([1.6, 2.05, 4.0, 9.5]) # The learning rate in the range 0.0-1.0 deeplearn.setLearningRate(0.5) # The percentage of dropouts in the range 0-100 deeplearn.setDropoutsPercent(0.001) # Title of the training error image deeplearn.setPlotTitle("White Wine Quality Training") # The number of time steps after which the training error image is redrawn deeplearn.setHistoryPlotInterval(500000) print "Training started" timeStep = 0 while (deeplearn.training() != 0): timeStep = timeStep + 1 print "Training Completed" print "Test data set performance is " + str(deeplearn.getPerformance()) + "%"; deeplearn.export("result.py") print "Exported trained network"
print str(noOfSamples) + " samples loaded" # The error threshold (percent) for each layer of the network. # Three hidden layers, plus the final training. # After going below the threshold the pre-training will move # on to the next layer deeplearn.setErrorThresholds([0.5, 0.5, 0.5, 2.0]) # The learning rate in the range 0.0-1.0 deeplearn.setLearningRate(0.2) # The percentage of dropouts in the range 0-100 deeplearn.setDropoutsPercent(0.001) # Title of the training error image deeplearn.setPlotTitle("Concrete Slump Training") # The number of time steps after which the training error image is redrawn deeplearn.setHistoryPlotInterval(900000) print "Training started" timeStep = 0 while (deeplearn.training() != 0): timeStep = timeStep + 1 print "Training Completed" print "Test data set performance is " + str(deeplearn.getPerformance()) + "%"; deeplearn.export("result.py") print "Exported trained network"