def main():
""" Driver.py """
upperCaseLettersInfoFile = 'info_1.csv'
greekLettersInfoFile = 'info_2.csv'
trainingData1File = 'train_1.csv'
trainingData2File = 'train_2.csv'
val1File = 'val_1.csv'
val2File = 'val_2.csv'
testWithLabel1File = 'test_with_label_1.csv'
testWithLabel2File = 'test_with_label_2.csv'
baseDTFile1 = 'Base-DT-DS1.csv'
baseDTFile2 = 'Base-DT-DS2.csv'
bestDTFile1 = 'Best-DT-DS1.csv'
bestDTFile2 = 'Best-DT-DS2.csv'
upperCaseLettersDict = utils.getInfo(upperCaseLettersInfoFile)
greekLettersDict = utils.getInfo(greekLettersInfoFile)
#Get data uses Pandas library. Returns 2d array with column headers in first row
trainingData1 = utils.getData(trainingData1File)
trainingData2 = utils.getData(trainingData2File)
#Generating our ML Models
baseDTUpperCase = decisionTree.generateBaseDT(trainingData1)
baseDTGreek = decisionTree.generateBaseDT(trainingData2)
bestDTUpperCase = decisionTree.generateBestDT(trainingData1)
bestDTGreek = decisionTree.generateBestDT(trainingData2)
val1Data = utils.getData(val1File)
val2Data = utils.getData(val2File)
testWithLabel1 = utils.getData(testWithLabel1File)
testWithLabel2 = utils.getData(testWithLabel2File)
utils.plotInstances(trainingData1, upperCaseLettersDict, 'Uppercase Letters', 'Training', 'trainingUppercase')
utils.plotInstances(trainingData2, greekLettersDict, 'Greek Letters', 'Training', 'trainingGreek')
utils.plotInstances(val1Data, upperCaseLettersDict, 'Uppercase Letters', 'Validation', 'validationUppercase')
utils.plotInstances(val2Data, greekLettersDict, 'Greek Letters', 'Validation', 'validationGreek')
utils.plotInstances(testWithLabel1, upperCaseLettersDict, 'Uppercase Letters', 'Test', 'testUppercase')
utils.plotInstances(testWithLabel2, greekLettersDict, 'Greek Letters', 'Test', 'testGreek')
print('Running Validation for Base DT - Upper Case Letters...')
utils.testModel(baseDTUpperCase, val1Data)
print('Running Validation for Best DT - Upper Case Letters...')
utils.testModel(bestDTUpperCase, val1Data)
print('\nRunning Tests for Base DT - Upper Case Letters...')
baseDTRes1 = utils.testModel(baseDTUpperCase, testWithLabel1)
utils.writeMLResults(baseDTRes1, baseDTFile1)
print('Running Tests for Best DT - Upper Case Letters...')
bestDTRes1 = utils.testModel(bestDTUpperCase, testWithLabel1)
utils.writeMLResults(bestDTRes1, bestDTFile1)
print('\nRunning Validation for Base DT - Greek Letters...')
utils.testModel(baseDTGreek, val2Data)
print('Running Validation for Best DT - Greek Letters...')
utils.testModel(bestDTGreek, val2Data)
print('\nRunning Tests for Base DT - Greek Letters...')
baseDTRes2 = utils.testModel(baseDTGreek, testWithLabel2)
utils.writeMLResults(baseDTRes2, baseDTFile2)
print('Running Tests for Best DT - Greek Letters...')
bestDTRes2 = utils.testModel(bestDTGreek, testWithLabel2)
utils.writeMLResults(bestDTRes2, bestDTFile2)
utils.plotConfusionMatrix(baseDTUpperCase, testWithLabel1, upperCaseLettersDict, 'Base DT Uppercase Letters' ,'Base-DT-DS1-CM')
utils.plotConfusionMatrix(baseDTGreek, testWithLabel2, greekLettersDict, 'Base DT Greek Letters' ,'Base-DT-DS2-CM')
utils.plotConfusionMatrix(bestDTUpperCase, testWithLabel1, upperCaseLettersDict, 'Best DT Uppercase Letters' ,'Best-DT-DS1-CM')
utils.plotConfusionMatrix(bestDTGreek, testWithLabel2, greekLettersDict, 'Best DT Greek Letters' ,'Best-DT-DS2-CM')
utils.getClassificationReport(baseDTUpperCase, testWithLabel1, upperCaseLettersDict, 'Base DT Uppercase Letters' ,'Base-DT-DS1-Report')
utils.getClassificationReport(baseDTGreek, testWithLabel2, greekLettersDict, 'Base DT Greek Letters' ,'Base-DT-DS2-Report')
utils.getClassificationReport(bestDTUpperCase, testWithLabel1, upperCaseLettersDict, 'Base DT Uppercase Letters' ,'Best-DT-DS1-Report')
utils.getClassificationReport(bestDTGreek, testWithLabel2, greekLettersDict, 'Base DT Greek Letters' ,'Best-DT-DS2-Report')
"score_test": high_score_test,
"accuracy_even": acc_even,
"accuracy_disc": acc_disc,
"accuracy_logr": acc_even
})
dffilename = os.path.join(result_path, "svm_joint.csv")
if not os.path.exists(os.path.dirname(dffilename)):
os.makedirs(os.path.dirname(dffilename))
df.to_csv(dffilename)
# Plot confusion matrix
target_names = ["intent"]*9
for i in range(9):
target_names[i] += str(i+1)
cnf_matrix = confusion_matrix(test_labels, high_prediction)
utils.plotConfusionMatrix(cnf_matrix, target_names)
# Plot trates accuracies
utils.plotAccBar(high_score_train, high_score_test, num_frames)
# Save predictions to files
# Save even weighted predictions
predevenfilename = os.path.join(result_path, "svm_joint_even.txt")
if not os.path.exists(os.path.dirname(predevenfilename)):
os.makedirs(os.path.dirname(predevenfilename))
np.savetxt(predevenfilename, pred_even, fmt="%d")
# Save discont weighted predictions
preddiscfilename = os.path.join(result_path, "svm_joint_disc.txt")
if not os.path.exists(os.path.dirname(preddiscfilename)):
os.makedirs(os.path.dirname(preddiscfilename))
np.savetxt(preddiscfilename, pred_disc, fmt="%d")
from sklearn.linear_model import LogisticRegression from sklearn.metrics import confusion_matrix from utils import plotConfusionMatrix import numpy as np import keras np.random.seed(1588390) # MARK: - load data (xTrain, yTrain), (xTest, yTest) = keras.datasets.mnist.load_data() xTrain = np.reshape(xTrain, (-1, 28*28)) / 255. xTest = np.reshape(xTest, (-1, 28*28)) / 255. # MARK: - logistic regression model = LogisticRegression(solver='saga', multi_class='multinomial', max_iter=100, verbose=2) model.fit(xTrain, yTrain) yPred = model.predict(xTest) confusionMatrix = confusion_matrix(yTest, yPred) plotConfusionMatrix(confusionMatrix, [i for i in range(10)], 'problem1.png')
###
model.evaluate(x_test, y_test)
val_loss = (history1.history['val_loss'])
train_loss = (history1.history['loss'])
plotLoss(val_loss, train_loss)
val_accuracy = history1.history['val_accuracy']
train_accuracy = history1.history['accuracy']
plotAccuracy(val_accuracy, train_accuracy)
y_pred = model.predict(x_test)
y_test_label = [class_labels[i] for i in np.argmax(y_test, axis=1)]
y_pred_label = [class_labels[i] for i in np.argmax(y_pred, axis=1)]
plotConfusionMatrix(class_labels,
y_test_label,
y_pred_label,
title='Confusion matrix - vgg16')
#---
### train model with data augmentation
opt2 = keras.optimizers.Adam(learning_rate=1e-5)
model_augmented = VggModel()
model_augmented.compile(optimizer=opt2,
loss='categorical_crossentropy',
metrics=['accuracy'])
model_checkpoint2 = ModelCheckpoint('vgg_model_augmented.h5',
'val_accuracy',
save_best_only=True)
my_callbacks2 = [model_checkpoint2, early_stop, reduce_lr]
history2 = model_augmented.fit(train_batches,
steps_per_epoch=len(x_train) // batch_size,
zScoreNormalization(trainInputs, testInputs)
# print(trainInputs[:5])
# print(testInputs[:5])
classifier = MLPClassifier(hidden_layer_sizes=(
5,
5,
),
activation='relu',
max_iter=1000,
solver='sgd',
verbose=10,
random_state=10,
learning_rate_init=.1)
classifier.fit(trainInputs, trainOutputs)
computedOutputs = classifier.predict(testInputs)
print("Computed: " + str(list(computedOutputs)))
print("Real: " + str(testOutputs))
accuracy, precision, recall, confusionMatrix = evalMultiClass(
testOutputs, list(computedOutputs), outputLabels)
print('accuracy: ', accuracy)
print('precision: ', precision)
print('recall: ', recall)
plotConfusionMatrix(confusionMatrix, outputLabels,
"Sepia vs Non-Sepia classification")
def main():
actionDatabaseDir = config.actionDatabaseDir
categories = config.actionCateogory
## Step.1 Data loading and features extraction
# Get features from training data over all categories
print "Data loading and feature extraction ..."
allFeatures = np.array([])
trainActionSequence = []
testActionSequence = []
if os.path.exists('allFeatures.npy') \
and os.path.exists('trainActionSequence.npy') \
and os.path.exists('testActionSequence.npy'):
allFeatures = np.load('allFeatures.npy')
trainActionSequence = np.load('trainActionSequence.npy')
testActionSequence = np.load('testActionSequence.npy')
else:
for category in categories:
categoryPath = os.path.join(actionDatabaseDir, category)
allData = os.listdir(categoryPath)
# Train data and test data loading
for data in allData:
filePath = os.path.join(categoryPath, data)
actionSequence = ActionSequence(filePath)
actionSequence.extractStip() # extract STIP
subject = actionSequence.subject
if subject in config.trainDataSubjects:
trainActionSequence.append(actionSequence)
if allFeatures.size == 0:
allFeatures = actionSequence.stipFeatures
else:
allFeatures = np.vstack((allFeatures,
actionSequence.stipFeatures))
elif subject in config.testDataSubjects:
testActionSequence.append(actionSequence)
np.save('allFeatures', allFeatures)
np.save('trainActionSequence', trainActionSequence)
np.save('testActionSequence', testActionSequence)
## Step.2 Codebook generation
print "Codebook generation ..."
bovw = BagOfWords(featureEncodingMethod = 'sparse-coding',
poolingMethod = 'max-pooling',
normalizationMethod = 'L2-norm')
# bovw = BagOfWords(featureEncodingMethod = 'vector-quantization',
# poolingMethod = 'sum-pooling',
# normalizationMethod = 'L1-norm')
if os.path.exists('codebook.npy'):
codebook = np.load('codebook.npy')
bovw.codebook = codebook
else:
bovw.generateCodebook(allFeatures)
np.save('codebook', bovw.codebook)
## Step.3 Feature encoding for train data
train_y = []
train_X = []
for actionSequence in trainActionSequence:
# Feature encoding, pooling, normalization
actionSequence.generateFinalFeatures(bovw)
# Format train data
train_y.append(actionSequence.categoryId)
train_X.append(actionSequence.finalFeatures)
# Cross validation
if config.is_cv:
# cross validation
crossValidate(train_y, train_X)
## Step.4 Classification
# Learning using SVM
svmTool = SvmTool()
print "Model learning ..."
svmTool.learnModel(train_y, train_X)
# Feature encoding for test data and classify data using learned model
print "Predicating ..."
numCorrect = 0
trueLabels = []
testPredLabels = []
removeall(config.predctDir)
for actionSequence in testActionSequence:
# Feature encoding, pooling, normalization
actionSequence.generateFinalFeatures(bovw)
# Format train data
test_y = [actionSequence.categoryId]
test_X = [actionSequence.finalFeatures]
p_label, _ = svmTool.doPredication(test_y, test_X)
predCagtegoryId = int(p_label[0])
predCategoryName = categories[predCagtegoryId]
# Write predicated action to predicated category
predCategoryPath = os.path.join(config.predctDir, predCategoryName)
if not os.path.exists(predCategoryPath):
os.makedirs(predCategoryPath)
predFilePath = os.path.join(predCategoryPath, actionSequence.filename)
f = open(predFilePath, 'w')
f.close()
if predCagtegoryId == actionSequence.categoryId:
numCorrect += 1
trueLabels.append(actionSequence.categoryName)
testPredLabels.append(predCategoryName)
# Calculate results
accuracy = numCorrect / float(len(testActionSequence))
print "accuracy: ", accuracy
# Plot confusion matrix
saveFilename = 'confusion_matrix.png'
plotConfusionMatrix(trueLabels, testPredLabels,
saveFilename, normalization = False)
saveFilename = 'confusion_matrix_norm.png'
plotConfusionMatrix(trueLabels, testPredLabels,
saveFilename, normalization = True)
history1 = pickle.load(open('base_history.pkl', 'rb'))
model.evaluate(x_test, y_test)
val_loss = (history1['val_loss'])
train_loss = (history1['loss'])
plotLoss(val_loss, train_loss)
val_accuracy = history1['val_accuracy']
train_accuracy = history1['accuracy']
plotAccuracy(val_accuracy, train_accuracy)
y_pred = model.predict(x_test)
y_test_label = [class_labels[i] for i in np.argmax(y_test, axis=1)]
y_pred_label = [class_labels[i] for i in np.argmax(y_pred, axis=1)]
plotConfusionMatrix(class_labels, y_test_label, y_pred_label)
###
model_augmented = keras.models.load_model('base_model_augmented.h5')
history2 = pickle.load(open('base_history_augmented.pkl', 'rb'))
model_augmented.evaluate(x_test, y_test)
val_loss = (history2['val_loss'])
train_loss = (history2['loss'])
plotLoss(val_loss, train_loss)
val_accuracy = history2['val_accuracy']
train_accuracy = history2['accuracy']
plotAccuracy(val_accuracy, train_accuracy)
