# Example of Naive Bayes classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
(X, d, Xt, dt) = loadData('G3') # load training and testing data
name = 'GaussianNB' # name of the classifier
params = '' # parameters of the classifier
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
PrintConfusion(dt, ds) # confusion matrix
PlotDecisionLines(clf, X) # decision lines
PlotFeatures(X, d, 'Naive Bayes') # feature space
# Example of Decision Tree classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
(X, d, Xt, dt) = loadData('G2') # load training and testing data
name = 'DecisionTreeClassifier' # name of the classifier
# parameters of the classifier
params = 'max_depth = 4, min_samples_leaf = 8,random_state = 0'
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
PrintConfusion(dt, ds) # confusion matrix
PlotDecisionLines(clf, X) # decision lines
PlotFeatures(X, d, 'Decision Tree') # feature space
# Example of Neural Network classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
'MLPClassifier', 'solver="adam", alpha=1e-5,hidden_layer_sizes=(3,2), random_state=1,max_iter=2000',
(X, d, Xt, dt) = loadData('G3') # load training and testing data
name = 'MLPClassifier' # name of the classifier
# parameters of the classifier
params = 'solver="adam", alpha=1e-5,hidden_layer_sizes=(7,2), random_state=1,max_iter=2000'
# params = 'solver="adam", alpha=1e-5,hidden_layer_sizes=(10,), random_state=1,max_iter=2000'
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
PrintConfusion(dt, ds) # confusion matrix
# PlotDecisionLines(clf,X) # decision lines
PlotFeatures(X, d, 'Neural Network') # feature space
# Example of SVM classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
(X, d, Xt, dt) = loadData('G3') # load training and testing data
name = 'SVC' # name of the classifier
# parameters of the classifier
params = 'kernel = "linear", gamma=0.2, C=0.1'
# params = 'kernel = "poly" , gamma=0.2, C=0.1, degree = 2'
# params = 'kernel = "rbf" , gamma=0.2,C=1'
# params = 'kernel = "sigmoid", gamma=0.01, C=0.01'
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
PrintConfusion(dt, ds) # confusion matrix
PlotDecisionLines(clf, X) # decision lines
PlotFeatures(X, d, 'SVM') # feature space
# Example of KNN classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
(X, d, Xt, dt) = loadData('G2') # load training and testing data
name = 'KNeighborsClassifier' # name of the classifier
params = 'n_neighbors=3' # parameters of the classifier
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
PrintConfusion(dt, ds) # confusion matrix
PlotDecisionLines(clf, X) # decision lines
PlotFeatures(X, d, 'KNN-3') # feature space
# Example of KNN classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
from sklearn.metrics import confusion_matrix, accuracy_score
import numpy as np
(X, d, Xt, dt) = loadData('G2') # load training and testing data
name = 'KNeighborsClassifier' # name of the classifier
K = 11 # number of neighbors
params = 'n_neighbors=' + str(K) # parameters of the classifier
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
d0 = testClassifier(clf, X) # classifier testing
acc = accuracy_score(dt, ds) # testing accuracy
acc0 = accuracy_score(d, d0) # training accuracy
Z = np.concatenate((X, Xt), axis=0) # all samples
PlotDecisionLines(clf, Z) # decision lines
PlotFeatures(X, d, 'Training (KNN-' + str(K) + ') Acc = ' +
str(acc0)) # feature space for training
PlotDecisionLines(clf, Z) # decision lines
PlotFeatures(Xt, dt, 'Testing (KNN-' + str(K) + ') Acc = ' +
str(acc)) # feature space for testing
# Example of dmin classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
(X, d, Xt, dt) = loadData('G2') # load training and testing data
name = 'NearestCentroid' # name of the classifier
params = '' # parameters of the classifier
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
PrintConfusion(dt, ds) # confusion matrix
PlotDecisionLines(clf, X) # decision lines
PlotFeatures(X, d, 'dmin') # feature space
# Example of AdaBoost classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
(X, d, Xt, dt) = loadData('G2') # load training and testing data
name = 'AdaBoostClassifier' # name of the classifier
params = 'n_estimators=100' # parameters of the classifier
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
PrintConfusion(dt, ds) # confusion matrix
PlotDecisionLines(clf, X) # decision lines
PlotFeatures(X, d, 'AdaBoost') # feature space
# Example of QDA classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
(X, d, Xt, dt) = loadData('G4') # load training and testing data
name = 'QuadraticDiscriminantAnalysis' # name of the classifier
params = '' # parameters of the classifier
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
PrintConfusion(dt, ds) # confusion matrix
PlotDecisionLines(clf, X) # decision lines
PlotFeatures(X, d, 'QDA') # feature space
# Example of Logistic Regression classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
(X, d, Xt, dt) = loadData('G2') # load training and testing data
name = 'LogisticRegression' # name of the classifier
params = 'C=0.1,solver="lbfgs"' # parameters of the classifier
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
PrintConfusion(dt, ds) # confusion matrix
PlotDecisionLines(clf, X) # decision lines
PlotFeatures(X, d, 'Logistic Regression') # feature space
# Example of Random Forest classifier
from clf_utils import loadData, defineClassifier, trainClassifier, testClassifier, PrintConfusion, PlotFeatures, PlotDecisionLines
(X, d, Xt, dt) = loadData('G2') # load training and testing data
name = 'RandomForestClassifier' # name of the classifier
params = 'n_estimators=20,random_state = 0' # parameters of the classifier
clf = defineClassifier([name, params]) # classifier definition
clf = trainClassifier(clf, X, d) # classifier training
ds = testClassifier(clf, Xt) # classifier testing
PrintConfusion(dt, ds) # confusion matrix
PlotDecisionLines(clf, X) # decision lines
PlotFeatures(X, d, 'Random Forest') # feature space