def createPerceptronClassifier(img, samples, class_names, n_samples, ops=None, filepath=None, params={}):
mp = MultilayerPerceptron()
if "learning_rate" in params:
# In (0, 1]
mp.setLearningRate(params.get("learning_rate", mp.getLearningRate()))
# Number of nodes per layer: a set of comma-separated values (numbers), or:
# 'a' = (number of attributes + number of classes) / 2
# 'i' = number of attributes,
# 'o' = number of classes
# 't' = number of attributes + number of classes.
# See MultilayerPerceptron.setHiddenLayers
# https://weka.sourceforge.io/doc.dev/weka/classifiers/functions/MultilayerPerceptron.html#setHiddenLayers-java.lang.String-
mp.setHiddenLayers(params.get("hidden_layers", "10,5"))
return trainClassifier(mp, img, samples, class_names, n_samples, ops=ops, filepath=filepath)
print "Usage: supervised.py <ARFF-file>"
sys.exit()
# load data file
print "Loading data..."
file = FileReader(sys.argv[1])
data = Instances(file)
# set the class Index - the index of the dependent variable
data.setClassIndex(data.numAttributes() - 1)
# define the algorithms to be used.
algo_list = [(NaiveBayes(), 'NaiveBayes'), (BayesNet(), 'BayesNet'),
(J48(), 'J48'), (JRip(), 'JRip'), (KStar(), 'KStar'),
(RandomForest(), 'RandomForest'), (AdaBoostM1(), 'AdaBoostM1'),
(MultilayerPerceptron(), 'MultilayerPerceptron'),
(LibSVM(), 'LibSVM')]
algo_dict = dict([(x[1], x[0]) for x in algo_list])
algo_keys = [
'NaiveBayes', 'J48', 'BayesNet', 'JRip', 'RandomForest', 'KStar',
'AdaBoostM1', 'LibSVM', 'MultilayerPerceptron'
]
# example to set kernal type on libsvm. Default is 2
#algo = algo_dict['LibSVM']
#tag = SelectedTag("1",algo.TAGS_KERNELTYPE) # 0 = linear, 1 = polynomial, 2 = radial basis function, 3 = sigmoid
#algo.setKernelType(tag)
# train classifiers but filter out the name column first
print "Training classifiers..."
for key in algo_keys:
file.write("epochs,rmse\n")
wallfile = "data/plot/" + str(os.path.splitext(os.path.basename(__file__))[0]) + "_" + \
str(os.path.splitext(os.path.basename(sys.argv[1]))[0]) + "_wall.csv"
filewall=open(wallfile, 'w', bufsize) # open a file for wall clock time
filewall.write("epochs,seconds\n")
logfile = "logs/" + str(os.path.splitext(os.path.basename(__file__))[0]) + "_" + \
str(os.path.splitext(os.path.basename(sys.argv[1]))[0]) + "_tunable.log"
log=open(logfile, 'w', bufsize) # open general log file
# loop for different number of training epochs
data.setClassIndex(data.numAttributes() - 1)
for num in range(1,1000,50):
log.write("---------------------------------\nEpoch: " + str(num) + "\n")
algo = MultilayerPerceptron()
algo.setTrainingTime(num)
x = time.time()
algo.buildClassifier(data)
log.write("Time to build classifier: " + str(time.time() - x) + "\n")
filewall.write(str(num) + "," + str(time.time() - x) + "\n")
evaluation = Evaluation(data)
output = PlainText() # plain text output for predictions
output.setHeader(data)
buffer = StringBuffer() # buffer to use
output.setBuffer(buffer)
attRange = Range() # no additional attributes output
outputDistribution = Boolean(False) # we don't want distribution
x = time.time()
#evaluation.evaluateModel(algo, data, [output, attRange, outputDistribution])
evaluation.crossValidateModel(algo, data, 10, rand, [output, attRange, outputDistribution])
filelimit.write("instances,pctincorrecttest,pctincorrecttrain\n")
logfile = "logs/" + classifiername + "_" + dataname + crossvalidate + ".log"
log=open(logfile, 'w', bufsize) # open general log file
timefilename = "data/plot/" + classifiername + "_" + dataname + crossvalidate + "_traintime.csv"
timefile = open(timefilename, 'w', bufsize)
timefile.write("instances,timetest,timetrain\n")
for num in range(int(p['mlp.initial']),fulltrainset.numInstances(),(fulltrainset.numInstances() / int(p['mlp.numdatapoints']))):
trainset = Instances(fulltrainset,0,num) # create training set
trainset.setClassIndex(trainset.numAttributes() - 1)
log.write("---------------------------------\nTraining Set Size: " + str(trainset.numInstances()) + ", Test Set Size: " + str(testset.numInstances()) + ", Full data set size: " + str(fulltrainset.numInstances()) + "\n")
filelimit.write(str(trainset.numInstances()))
timefile.write(str(num))
for dataset in [testset, fulltrainset]:
algo = MultilayerPerceptron()
algo.setTrainingTime(int(p['mlp.N']))
x = time.time()
algo.buildClassifier(trainset)
evaluation = Evaluation(trainset)
timefile.write("," + str(time.time() - x))
output = PlainText() # plain text output for predictions
output.setHeader(trainset)
buffer = StringBuffer() # buffer to use
output.setBuffer(buffer)
attRange = Range() # no additional attributes output
outputDistribution = Boolean(False) # we don't want distribution
x = time.time()
if (int(crossvalidate)):
evaluation.crossValidateModel(algo, dataset, 10, rand, [output, attRange, outputDistribution])
else:
timefile = open(timefilename, 'w', bufsize)
timefile.write("instances,timetest,timetrain\n")
for num in range(int(p['mlp.initial']), fulltrainset.numInstances(),
(fulltrainset.numInstances() / int(p['mlp.numdatapoints']))):
trainset = Instances(fulltrainset, 0, num) # create training set
trainset.setClassIndex(trainset.numAttributes() - 1)
log.write("---------------------------------\nTraining Set Size: " +
str(trainset.numInstances()) + ", Test Set Size: " +
str(testset.numInstances()) + ", Full data set size: " +
str(fulltrainset.numInstances()) + "\n")
filelimit.write(str(trainset.numInstances()))
timefile.write(str(num))
for dataset in [testset, fulltrainset]:
algo = MultilayerPerceptron()
algo.setTrainingTime(int(p['mlp.N']))
x = time.time()
algo.buildClassifier(trainset)
evaluation = Evaluation(trainset)
timefile.write("," + str(time.time() - x))
output = PlainText() # plain text output for predictions
output.setHeader(trainset)
buffer = StringBuffer() # buffer to use
output.setBuffer(buffer)
attRange = Range() # no additional attributes output
outputDistribution = Boolean(False) # we don't want distribution
x = time.time()
if (int(crossvalidate)):
evaluation.crossValidateModel(
algo, dataset, 10, rand,
# check commandline parameters
if (not (len(sys.argv) == 3)):
print "Usage: weka.py <ARFF-file>"
sys.exit()
file = FileReader(sys.argv[1])
file2 = FileReader(sys.argv[2])
data = Instances(file)
test = Instances(file2)
data.setClassIndex(data.numAttributes() - 1)
test.setClassIndex(test.numAttributes() - 1)
evaluation = Evaluation(data)
buffer = StringBuffer()
attRange = Range() # no additional attributes output
outputDistribution = Boolean(False) # we don't want distribution
nn = MultilayerPerceptron()
nn.buildClassifier(data) # only a trained classifier can be evaluated
#print evaluation.evaluateModel(nn, ['-t', sys.argv[1], '-T', sys.argv[2]])#;, [buffer, attRange, outputDistribution])
res = evaluation.evaluateModel(nn, test,
[buffer, attRange, outputDistribution])
f = open('predictions/' + data.relationName(), 'w')
for d in res:
f.write(str(d) + '\n')
f.close()
SerializationHelper.write("models/" + data.relationName() + ".model", nn)
# print out the built model
#print "--> Generated model:\n"
#print nn
if (not (len(sys.argv) == 2)):
print "Usage: supervised.py <ARFF-file>"
sys.exit()
# load data file
print "Loading data..."
file = FileReader(sys.argv[1])
data = Instances(file)
# set the class Index - the index of the dependent variable
data.setClassIndex(data.numAttributes() - 1)
# define the algorithms to be used.
algo_list = [(NaiveBayes(), 'NaiveBayes'), (BayesNet(),'BayesNet'), (J48(),'J48'), (JRip(), 'JRip'),
(KStar(), 'KStar'), (RandomForest(), 'RandomForest'), (AdaBoostM1(),'AdaBoostM1'),
(MultilayerPerceptron(),'MultilayerPerceptron'), (LibSVM(), 'LibSVM')]
algo_dict = dict([(x[1], x[0]) for x in algo_list])
algo_keys = ['NaiveBayes', 'J48', 'BayesNet', 'JRip', 'RandomForest', 'KStar', 'AdaBoostM1', 'LibSVM', 'MultilayerPerceptron']
# example to set kernal type on libsvm. Default is 2
algo = algo_dict['LibSVM']
tag = SelectedTag("1",algo.TAGS_KERNELTYPE) # 0 = linear, 1 = polynomial, 2 = radial basis function, 3 = sigmoid
algo.setKernelType(tag)
# train classifiers
print "Training classifiers..."
for key in algo_keys :
algo = algo_dict[key]
algo.buildClassifier(data)
# evaluate classifiers and print a result summary including confusion matrix
# check commandline parameters
if (not (len(sys.argv) == 3)):
print "Usage: weka.py <ARFF-file>"
sys.exit()
file = FileReader(sys.argv[1])
file2 = FileReader(sys.argv[2])
data = Instances(file)
test = Instances(file2)
data.setClassIndex(data.numAttributes() - 1)
test.setClassIndex(test.numAttributes() - 1)
evaluation = Evaluation(data)
buffer = StringBuffer()
attRange = Range() # no additional attributes output
outputDistribution = Boolean(False) # we don't want distribution
nn = MultilayerPerceptron()
nn.buildClassifier(data) # only a trained classifier can be evaluated
#print evaluation.evaluateModel(nn, ['-t', sys.argv[1], '-T', sys.argv[2]])#;, [buffer, attRange, outputDistribution])
res = evaluation.evaluateModel(nn, test, [buffer, attRange, outputDistribution])
f = open('predictions/' + data.relationName(), 'w')
for d in res:
f.write(str(d) + '\n');
f.close()
SerializationHelper.write("models/" + data.relationName() + ".model", nn)
# print out the built model
#print "--> Generated model:\n"
#print nn
for i in range(len(variances)):
stack.addSlice("Variance" + str(variances[i]), imgvars[i].getProcessor())
for i in range(len(channels)):
stack.addSlice("channel" + str(i + 1), channels[i].getProcessor())
# create empty feature stack
features = FeatureStack(stack.getWidth(), stack.getHeight(), False)
# set my features to the feature stack
features.setStack(stack)
# put my feature stack into the array
featuresArray.set(features, 0)
featuresArray.setEnabledFeatures(features.getEnabledFeatures())
mp = MultilayerPerceptron()
hidden_layers = "%i,%i,%i" % (20, 14, 8)
mp.setHiddenLayers(hidden_layers)
mp.setLearningRate(0.7)
mp.setDecay(True)
mp.setTrainingTime(200)
mp.setMomentum(0.3)
wekaSegmentation = WekaSegmentation(image)
wekaSegmentation.setFeatureStackArray(featuresArray)
wekaSegmentation.setClassifier(mp)
wekaSegmentation.addExample(0, posroi, 1)
wekaSegmentation.addExample(1, negroi, 1)
wekaSegmentation.trainClassifier()
wekaSegmentation.saveClassifier(folder + "\\vessel-classifier_big.model")
