Assignment
In the second ML assignment you have to compare theperformance of three different classification algorithms, namely Naive Bayes,SVM, and Random Forest. For this assignment you need to generate a randombinary classification problem, and then train and test (using 10-fold crossvalidation) the three algorithms. For some algorithms inner cross validation(5-fold) for choosing the parameters is needed. Then, show the classificationperformace (per-fold and averaged) in the report, and briefly discussing the results.
Steps
Create a classification dataset (n samples 1000, nfeatures 10)
Split the dataset using 10-fold cross validation
Train the algorithms
GaussianNB
SVC (possible C values [1e-02, 1e-01, 1e00, 1e01, 1e02],RBF kernel)
RandomForestClassifier (possible n estimators values [10,100, 1000])
Evaluate the cross-validated performance
Accuracy
F1-score
AUC ROC
代码:
from sklearn import datasets
from sklearn import cross_validation
from sklearn.naive_bayes import GaussianNB
from sklearn.svm import SVC
from sklearn.ensemble import RandomForestClassifier
from sklearn import metrics
dataset = datasets.make_classification(n_samples=2000, n_features=10,
n_informative=2, n_redundant=2, n_repeated=0, n_classes=2)
# spilt using 10-fold
kf = cross_validation.KFold(1000, n_folds=10, shuffle=True)
for train_index, test_index in kf:
X_train, y_train = dataset[0][train_index], dataset[1][train_index]
X_test, y_test = dataset[0][test_index], dataset[1][test_index]
# Gaussian Naive Bayes
clf = GaussianNB()
clf.fit(X_train, y_train)
pred = clf.predict(X_test)
print("GaussianNB:")
print("pred: \n", pred)
print("y_test: \n", y_test)
# Evaluate the cross-validated performance
print("Evaluate the cross-validated performance:")
acc = metrics.accuracy_score(y_test, pred)
print("Accuracy: ", acc)
f1 = metrics.f1_score(y_test, pred)
print("F1-score: ",f1)
auc = metrics.roc_auc_score(y_test, pred)
print("AUC ROC: ", auc)
print("\n")
# SVC
clf = SVC(C=1e-01, kernel='rbf', gamma=0.1)
clf.fit(X_train, y_train)
pred = clf.predict(X_test)
print("SVC: ")
print("pred: \n", pred)
print("y_test: \n", y_test)
# Evaluate the cross-validated performance
print("Evaluate the cross-validated performance:")
acc = metrics.accuracy_score(y_test, pred)
print("Accuracy: ", acc)
f1 = metrics.f1_score(y_test, pred)
print("F1-score: ",f1)
auc = metrics.roc_auc_score(y_test, pred)
print("AUC ROC: ", auc)
print("\n")
# Random Forest
clf = RandomForestClassifier(n_estimators=6)
clf.fit(X_train, y_train)
pred = clf.predict(X_test)
print("RandomForestClassifier: ")
print("pred: \n", pred)
print("y_test: \n", y_test)
# Evaluate the cross-validated performance
print("Evaluate the cross-validated performance:")
acc = metrics.accuracy_score(y_test, pred)
print("Accuracy: ", acc)
f1 = metrics.f1_score(y_test, pred)
print("F1-score: ",f1)
auc = metrics.roc_auc_score(y_test, pred)
print("AUC ROC: ", auc)
print("\n")
运行的结果:
GaussianNB:
pred:
[0 0 1 1 0 0 0 1 0 0 0 1 0 0 0 1 1 0 0 0 0 0 0 1 0 0 1 0 1 0 1 0 1 1 1 0 0
1 1 1 1 0 1 1 0 0 1 1 0 1 1 0 0 0 1 0 1 0 0 0 1 1 1 1 1 0 0 1 1 1 1 0 0 0
1 1 1 0 0 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1 0 0 0 0 0]
y_test:
[0 0 1 1 1 0 0 1 0 1 0 1 0 0 0 1 1 0 0 0 0 0 0 1 0 1 1 1 1 0 0 0 1 1 1 0 0
0 1 1 1 0 1 1 0 0 1 1 0 1 1 0 0 0 1 0 1 0 0 0 1 1 1 0 1 0 0 1 1 1 1 0 0 1
1 1 1 0 0 1 1 0 1 1 0 1 1 0 1 1 1 1 0 1 1 0 0 1 0 0]
Evaluate the cross-validated performance:
Accuracy: 0.88
F1-score: 0.8799999999999999
AUC ROC: 0.8814102564102564
SVC:
pred:
[0 0 1 1 0 0 0 1 0 0 0 1 0 0 0 1 1 0 0 0 0 0 0 1 0 0 1 0 1 0 0 0 1 1 1 0 0
0 1 1 1 0 1 1 0 0 1 1 0 1 1 0 0 0 1 0 1 0 0 0 1 1 1 1 1 0 0 1 1 1 1 0 0 0
1 1 1 0 0 1 1 0 1 1 0 1 0 1 1 0 1 1 0 1 1 0 0 0 0 0]
y_test:
[0 0 1 1 1 0 0 1 0 1 0 1 0 0 0 1 1 0 0 0 0 0 0 1 0 1 1 1 1 0 0 0 1 1 1 0 0
0 1 1 1 0 1 1 0 0 1 1 0 1 1 0 0 0 1 0 1 0 0 0 1 1 1 0 1 0 0 1 1 1 1 0 0 1
1 1 1 0 0 1 1 0 1 1 0 1 1 0 1 1 1 1 0 1 1 0 0 1 0 0]
Evaluate the cross-validated performance:
Accuracy: 0.9
F1-score: 0.8979591836734695
AUC ROC: 0.9022435897435899
RandomForestClassifier:
pred:
[0 0 1 1 0 0 0 1 0 0 0 1 0 0 0 1 1 0 0 0 0 0 0 1 0 0 1 0 1 0 0 0 1 1 1 0 0
0 1 1 1 0 1 1 0 0 1 1 0 1 1 0 0 0 1 0 1 0 0 0 1 1 1 1 1 0 0 1 1 1 1 0 0 1
1 1 1 0 0 1 1 0 1 1 0 1 0 0 1 1 1 1 0 1 1 0 0 1 0 0]
y_test:
[0 0 1 1 1 0 0 1 0 1 0 1 0 0 0 1 1 0 0 0 0 0 0 1 0 1 1 1 1 0 0 0 1 1 1 0 0
0 1 1 1 0 1 1 0 0 1 1 0 1 1 0 0 0 1 0 1 0 0 0 1 1 1 0 1 0 0 1 1 1 1 0 0 1
1 1 1 0 0 1 1 0 1 1 0 1 1 0 1 1 1 1 0 1 1 0 0 1 0 0]
Evaluate the cross-validated performance:
Accuracy: 0.94
F1-score: 0.9400000000000001
AUC ROC: 0.9415064102564101
由三组性能结果可以看出,在这组数据时,RandomForestClassifier性能在Accuracy,F1-score,AUC ROC中都较其他两种算法要好。