def runClassifier(self, X_train, Y_train, X_test, Y_test):
classalgs = {
'Logistic_Regression': algs.LogisticRegression1(),
'Gauassian_SVM': algs.SVM1(),
'Logistic_Regression_weighted ': algs.LogisticRegression1wt(),
'Gauassian_SVM_weighted': algs.SVM1wt()
}
for learnername, learner in classalgs.iteritems():
print 'Running learner = ' + learnername
learner.learn(X_train, Y_train)
predictions = learner.predict(X_test)
recall = util.getRecall(Y_test, predictions)
print '\n Recall for ' + learnername + ': ' + str(recall)
precision = util.getPrecision(Y_test, predictions)
print '\n Precision for ' + learnername + ': ' + str(precision)
f5_score = util.getF5(precision, recall)
print '\n F5 Score for ' + learnername + ': ' + str(f5_score)
AUCROC = util.getAUCROC(Y_test, predictions)
print '\n AUC ROC Score for ' + learnername + ': ' + str(AUCROC)
AUCROCPlotPoints = util.getAUCROCPlotPoints(Y_test, predictions)
print('\n tpr : {0} fpr : {1} auc_roc : {2} learnername : {3}'
).format(AUCROCPlotPoints[0], AUCROCPlotPoints[1],
AUCROCPlotPoints[2], learnername)
self.plotGraph(AUCROCPlotPoints, learnername)
f1_score = util.fscore(Y_test, predictions)
print '\n f1_score for ' + learnername + ': ' + str(f1_score)
accuracy = util.getaccuracy(Y_test, predictions)
print 'Accuracy for ' + learnername + ': ' + str(accuracy)
def runClassifier(self,X_train,Y_train,X_test,Y_test):
classalgs = {
'Logistic_Regression' : algs.LogisticRegression1(),
'Gauassian_SVM' :algs.SVM1(),
'Logistic_Regression_weighted ' : algs.LogisticRegression1wt(),
'Gauassian_SVM_weighted' :algs.SVM1wt()
}
for learnername, learner in classalgs.iteritems():
print 'Running learner = ' + learnername
learner.learn(X_train, Y_train)
predictions = learner.predict(X_test)
recall = util.getRecall(Y_test, predictions)
print '\n Recall for ' + learnername + ': ' + str(recall)
precision=util.getPrecision(Y_test, predictions)
print '\n Precision for ' + learnername + ': ' + str(precision)
f5_score=util.getF5(precision,recall)
print '\n F5 Score for ' + learnername + ': ' + str(f5_score)
AUCROC =util.getAUCROC(Y_test, predictions)
print '\n AUC ROC Score for ' + learnername + ': ' + str(AUCROC)
AUCROCPlotPoints =util.getAUCROCPlotPoints(Y_test, predictions)
print('\n tpr : {0} fpr : {1} auc_roc : {2} learnername : {3}').format(AUCROCPlotPoints[0],AUCROCPlotPoints[1],AUCROCPlotPoints[2],learnername)
self.plotGraph(AUCROCPlotPoints,learnername)
f1_score = util.fscore(Y_test, predictions)
print '\n f1_score for ' + learnername + ': ' + str(f1_score)
accuracy = util.getaccuracy(Y_test, predictions)
print 'Accuracy for ' + learnername + ': ' + str(accuracy)
#'Neural Net_64':algs.backPropogation((Xtrain.shape[1],64,1))
}
accuracyD={}
for learnername, learner in classalgs.iteritems():
print 'Running learner = ' + learnername
learner.learn(Xtrain, Ytrain)
predictions = learner.predict(Xtest)
recall = util.getRecall(Ytest, predictions)
print '\n Recall for ' + learnername + ': ' + str(recall)
AUCROC =util.getAUCROC(Ytest, predictions)
print '\n AUC ROC Score for ' + learnername + ': ' + str(AUCROC)
AUCROCPlotPoints =util.getAUCROCPlotPoints(Ytest, predictions)
print('\n tpr : {0} fpr : {1} auc_roc : {2} learnername : {3}').format(AUCROCPlotPoints[0],AUCROCPlotPoints[1],AUCROCPlotPoints[2],learnername)
self.plotGraph(AUCROCPlotPoints,learnername)
f1_score = util.fscore(Ytest, predictions)
print '\n f1_score for ' + learnername + ': ' + str(f1_score)
accuracy = util.getaccuracy(Ytest, predictions)
print 'Accuracy for ' + learnername + ': ' + str(accuracy)
accuracyD[learnername]=AUCROC
FoldAccuracy[i]=accuracyD
i=i+1
self.StatisticalSignificance(FoldAccuracy)
def plotGraph(self,AUCROCPlotPoints,learnerName):
'''
Plot the AUC-ROC plot for each classifier.
'''
fpr=AUCROCPlotPoints[0]
tpr=AUCROCPlotPoints[1]
roc_auc= metrics.auc(fpr, tpr)