def train_rbf_svm(X, y):
svc = SVC(kernel='rbf', gamma=1.0, C=1.0, random_state=1)
svc.fit(X, y)
plot_decision_regions(X=X,
y=y,
classifier=svc,
test_idx=None,
xlabel='x',
ylabel='y',
title='SVC')
return svc
def train_svc(X_train_std, y_train, X_test_std, y_test):
X_combined_std = np.vstack((X_train_std, X_test_std))
y_combined = np.hstack((y_train, y_test))
svc = SVC(kernel='linear', C=100.0, random_state=1)
svc.fit(X_train_std, y_train)
y_pred = svc.predict(X_test_std)
print("Accuracy: %.2f" % (accuracy_score(y_test, y_pred)))
plot_decision_regions(X=X_combined_std,
y=y_combined,
classifier=svc,
test_idx=range(105, 150),
xlabel='petal length [standardized]',
ylabel='petal width [standardized]',
title='SVC')
return svc
def train_logistic(X_train_std, y_train, X_test_std, y_test):
X_combined_std = np.vstack((X_train_std, X_test_std))
y_combined = np.hstack((y_train, y_test))
lgr = LogisticRegression(C=100.0, random_state=1)
lgr.fit(X_train_std, y_train)
y_pred = lgr.predict(X_test_std)
print("Accuracy: %.2f" % (accuracy_score(y_test, y_pred)))
plot_decision_regions(X=X_combined_std,
y=y_combined,
classifier=lgr,
test_idx=range(105, 150),
xlabel='petal length [standardized]',
ylabel='petal width [standardized]',
title='Logistic')
return lgr
def train_logisticBGD(X, y):
lrgd = LogisticBGD(eta=0.05, n_iter=20, random_state=1)
lrgd.fit(X, y)
plot_decision_regions(X=X,
y=y,
classifier=lrgd,
test_idx=None,
xlabel='petal length [standardized]',
ylabel='petal width [standardized]',
title='LogisticBGD')
plt.plot(range(1, len(lrgd.costs) + 1), lrgd.costs, marker='o')
plt.xlabel('epoch')
plt.xlabel('cost')
plt.xlabel('LogisticBGD-costs')
plt.show()
def train_perceptron(X_train_std, y_train, X_test_std, y_test):
ppn = Perceptron(max_iter=50,
tol=None,
eta0=0.1,
shuffle=True,
random_state=1,
verbose=1)
ppn.fit(X_train_std, y_train)
y_pred = ppn.predict(X_test_std)
print("Actual iter: %d" % ppn.n_iter_)
print("Misclassified samples: %d of %d" %
((y_test != y_pred).sum(), y_test.sum()))
print("Accuracy: %.2f" % (accuracy_score(y_test, y_pred)))
X_combined_std = np.vstack((X_train_std, X_test_std))
y_combined = np.hstack((y_train, y_test)) # y have only 1 dim
plot_decision_regions(X=X_combined_std,
y=y_combined,
classifier=ppn,
test_idx=range(105, 150),
xlabel='petal length [standardized]',
ylabel='petal width [standardized]',
title='Perceptron')
return ppn
y = df.iloc[0:100, 4].values
#convert label to int
#for logidstic classification
y = np.where(y == 'Iris-setosa', 0, 1)
print(y)
#iloc designate
x = df.iloc[0:100, [0, 2]].values
# deep copy x
x_std = np.copy(x)
#standalization
x_std[:,0] = (x_std[:,0] - x_std[:,0].mean())/x_std[:,0].std()
x_std[:,1] = (x_std[:,1] - x_std[:,1].mean())/x_std[:,1].std()
#Training
clf = LogisticRegressionGD(n_iter=100, eta=0.01, random_state=1).fit(x_std,y)
iris.plot_decision_regions(x_std,y,classifier=clf)
plt.title('AdalineGD')
plt.xlabel('sepal length standarlized')
plt.ylabel('petal length standarlized')
plt.legend(loc='upper left')
plt.tight_layout()
plt.show()
plt.plot(range(1,len(clf.cost_)+1),clf.cost_, marker='o')
plt.xlabel('Epochs')
plt.ylabel('Sum Squeared Error')
plt.tight_layout()
plt.show()
"""
fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(10, 4))
self.errors.append(error)
return self
def net_input(self, X):
return np.dot(X, self.weights[1:]) + self.weights[0]
def predict(self, X):
return np.where(self.net_input(X) >= 0.0, 1, -1)
def train_perceptron(X, y, verbose=True):
ppn = Perceptron(eta=0.1, n_iter=20)
ppn.fit(X, y)
if verbose:
plt.plot(range(1, len(ppn.errors) + 1), ppn.errors, marker='o')
plt.xlabel('epoch')
plt.ylabel('error')
plt.show()
return ppn
if __name__ == "__main__":
X, y = load_iris_data()
ppn = train_perceptron(X, y, verbose=False)
plot_decision_regions(X,
y,
ppn,
xlabel="sepal length [cm]",
ylabel="petal length [cm]",
title="perceptron")