def knn_maker_algoritm(trainfolderpath):
dataModel = get_data_from_folder(trainfolderpath)
knnModel = classify(dataModel.trainFeat, dataModel.trainLabels)
accKnnModel = getAccuracy(knnModel, dataModel.testFeat,
dataModel.testLabels)
print("KNNClassifier accuracy: {:.2f}%".format(accKnnModel * 100))
saveModel(knnModel)
def mlp_maker_algoritm(trainfolderpath):
dataModel = get_data_from_folder(trainfolderpath)
mlpModel = classify(dataModel.trainFeat, dataModel.trainLabels)
accmlpModel = getAccuracy(mlpModel, dataModel.testFeat,
dataModel.testLabels)
print("MLP Classifier accuracy: {:.2f}%".format(accmlpModel * 100))
saveModel(mlpModel)
def random_forest_maker_algoritm(trainfolderpath):
dataModel = get_data_from_folder(trainfolderpath)
random_forestModel = classify(dataModel.trainFeat, dataModel.trainLabels)
accModel = getAccuracy(random_forestModel, dataModel.testFeat,
dataModel.testLabels)
print("Random Forest Classifier accuracy: {:.2f}%".format(accModel * 100))
saveModel(random_forestModel)
def decision_tree_maker_algoritm(trainfolderpath):
dataModel = get_data_from_folder(trainfolderpath)
decisionTreeModel = classify(dataModel.trainFeat, dataModel.trainLabels)
accDecitionTree = getAccuracy(decisionTreeModel, dataModel.testFeat,
dataModel.testLabels)
print("DecisionTreeClassifier accuracy: {:.2f}%".format(accDecitionTree *
100))
saveModel(decisionTreeModel)
def chooseClassifier(trainfolderpath):
dataModel = get_data_from_folder(trainfolderpath)
decisionTreeModel = dtc(dataModel.trainFeat, dataModel.trainLabels)
knnModel = kc(dataModel.trainFeat, dataModel.trainLabels)
accDecitionTree = getAccuracy(decisionTreeModel, dataModel.testFeat,
dataModel.testLabels)
accKnnModel = getAccuracy(knnModel, dataModel.testFeat,
dataModel.testLabels)
print("DecisionTreeClassifier accuracy: {:.2f}%".format(accDecitionTree *
100))
print("KNNClassifier accuracy: {:.2f}%".format(accKnnModel * 100))
acc = accDecitionTree
if accKnnModel > acc:
acc = accKnnModel
print("Knn is saved")
saveModel(knnModel)
else:
print("Decition tree is saved")
saveModel(decisionTreeModel)
return acc
def logistic_regression_maker_algoritm(trainfolderpath):
dataModel = get_data_from_folder(trainfolderpath)
model = classify(dataModel.trainFeat, dataModel.trainLabels)
acc = getAccuracy(model, dataModel.testFeat, dataModel.testLabels)
print("Logistic Regression accuracy: {:.2f}%".format(acc * 100))
saveModel(model)
def svm_maker_algoritm(trainfolderpath):
dataModel = get_data_from_folder(trainfolderpath)
svmModel = classify(dataModel.trainFeat, dataModel.trainLabels)
accSVMModel = getAccuracy(svmModel, dataModel.testFeat, dataModel.testLabels)
print("SVMClassifier accuracy: {:.2f}%".format(accSVMModel * 100))
saveModel(svmModel)
loadedUp = False
while not loadedUp:
try:
trainImages = helper.loadData(trainNP+"XtrainImages")
trainVecs = helper.loadData(trainNP+"XtrainVecs")
trainTargets = helper.loadData(trainNP+"ytrain")
loadedUp = True
except Exception as e:
err = e
print err, " \r",
time.sleep(2)
print ""
subprocess.call("rm "+trainNP+"XtrainImages.h5",shell=True)
subprocess.call("rm "+trainNP+"XtrainVecs.h5",shell=True)
subprocess.call("rm "+trainNP+"ytrain.h5",shell=True)
#train the model on it
print trainImages.shape, trainVecs.shape
model.fit([trainImages,trainVecs], trainTargets, batch_size=batch_size, nb_epoch=1)
del trainImages, trainTargets
helper.saveModel(model,folder+"wholeModel")