def main():
data = pd.read_csv('data/iris.data', header=None)
"""
columns in iris.data are as following
sepal length | sepal width | petal length | petal width | species
first we will show how our data looks like
"""
# print('test how long it is:{}'.format(data.iloc[:, :].values.shape))
y = data.iloc[
0:100,
4].values # labels of data ( species column), we take first 100 to work only on 2 classes
y = np.where(y == 'Iris-setosa', -1,
1) # now were Iris-setosa we set class to -1, else class is 1
X = data.iloc[0:100, [0, 2]].values # extracted sepal and petal length
plt.scatter(X[:50, 0],
X[:50, 1],
color='red',
marker='o',
label='iris-setosa')
plt.scatter(X[50:, 0],
X[50:, 1],
color='green',
marker='x',
label='iris-versicolor')
plt.title('iris-setosa and iris-versicolor')
plt.xlabel('sepal length[cm]')
plt.ylabel('petal length[cm]')
plt.legend(loc='upper left')
plt.tight_layout()
plt.draw()
plt.waitforbuttonpress()
my_prcpt = Perceptron(eta=0.1, n_iter=10)
my_prcpt.fit(X,
y) # fitting with sepal and petal length (X) and classes (y)
# now present how many errors where made when learning during each data traversal
plt.figure()
plt.plot(range(1, len(my_prcpt.errors_) + 1), my_prcpt.errors_, marker='.')
plt.xticks(np.arange(1, len(my_prcpt.errors_) + 1, 1.0))
plt.title('Errors made in following epochs')
plt.xlabel('epochs')
plt.ylabel('errors in prediction')
plt.tight_layout()
plt.draw()
plt.waitforbuttonpress()
plt.figure()
plot_decision_regions(X, y, classifier=my_prcpt)
plt.title('Decision regions')
plt.xlabel('sepal length [cm]')
plt.ylabel('petal length [cm]')
plt.legend(loc='upper left')
plt.tight_layout()
plt.draw()
plt.waitforbuttonpress()
def run_average_face_perceptron(rate, epochs, standardize=False):
# extract just the face images from both the training and validation datasets
faces = feature_extraction.get_face_images(train_and_valid_images, train_and_valid_labels)
# find the average of all the face images and subtract is from every image
# this also centers all of the images
new_training_images = feature_extraction.find_average_face(faces, train_images)
# Find the average image in the testing set and subtract it from every image
# this allows us to compare the testing images more accurately
new_test_images = feature_extraction.find_average_face(test_images, test_labels)
if standardize:
standardized_train_images = StandardScaler().fit_transform(new_training_images)
standardized_test_images = StandardScaler().fit_transform(new_test_images)
average_face_perceptron = Perceptron(rate, epochs, "Perceptron Average Face (Standardized)")
average_face_perceptron.fit(standardized_train_images, train_labels)
average_results = average_face_perceptron.predict(standardized_test_images)
average_face_perceptron.calculate_results(average_results, test_labels, 150)
average_face_perceptron.graph_perceptron()
else:
average_face_perceptron = Perceptron(rate, epochs, "Perceptron Average Face")
average_face_perceptron.fit(new_training_images, train_labels)
average_results = average_face_perceptron.predict(new_test_images)
average_face_perceptron.calculate_results(average_results, test_labels, 150)
average_face_perceptron.graph_perceptron()
def main():
# chargement des données
# le fichier .csv contient 3 groupes de points 2D
# la première colonne du fichier correspond à x1, la deuxième à x2
# et la troisième correspond au groupe auquel est associé le point
filepath = "./data/test.csv"
data, labels = load_dataset(filepath)
# On garde le groupe de points 1 et 2
data = data[(labels == 0) | (labels == 1)]
labels = labels[(labels == 0) | (labels == 1)]
labels = np.where(labels == 1, 1, -1)
# On garde le groupe de points 1 et 3
# data = data[(labels==0) | (labels==2)]
# labels = labels[(labels==0) | (labels==2)]
# labels = np.where(labels == 0, 1, -1)
# Instanciation de la classe perceptron
p = Perceptron(2,
learning_rate=0.2,
lr_decay=False,
early_stopping=True,
display=True)
# Apprentissage
p.fit(data, labels)
# Score
#score = p.score(data, labels)
#print("precision : {:.2f}".format(score))
input("Press any key to exit...")
def main():
# load data into shuffle matrix
df = pd.read_csv("diabetes.csv")
df = np.array(df)
np.random.shuffle(df)
# split data into labels and design matrix
design = df[:, :-1].T
labels = df[:, -1]
labels = labels.reshape((labels.shape[0],1)).T # keep dimensionality to 2
_, m_tot = design.shape
# split into test and training data
frac_test = .6
split_idx = int(frac_test*m_tot)
train_design = design[:, :split_idx]
train_labels = labels[:, :split_idx]
test_design = design[:, split_idx:]
test_labels = labels[:, split_idx:]
# fit perceptron
perc = Perceptron()
perc.fit(X=train_design, Y=train_labels, alpha=1e-4,
lambd=0, epochs=100_000)
# get model accuracies
test_acc = perc.acc(X=test_design, Y=test_labels)
train_acc = perc.acc(X=train_design, Y=train_labels)
print("Test set accuracy: %.5f" % test_acc)
print("Training set accuracy: %.5f" % train_acc)
def classification(self):
ppt = Perceptron(eta=0.1, n_iter=10)
ppt.fit(self.x, self.y)
plt.plot(range(1, len(ppt.errors_) + 1), ppt.errors_, marker="o")
plt.savefig("miss_classification.png")
def main():
#load dataset
data = pd.read_csv('../data/iris.data', sep=',', header=0)
#classifying with two classes only
dataset = data[data['CLASS'] != 'Iris-virginica']
#using the LENGHTS for separation
X = dataset[['SEPAL_LENGHT','PETAL_LENGHT']]
#converting classes to -1 and 1
y = dataset['CLASS'].apply(convert)
#train Perceptron
model = Perceptron(X.shape[1])
model.fit(X,y)
#plotting
w = model.get_w()
#decision boundary: w[0] + w[1]*x + w[2]*y = 0
x1 = np.linspace(0, 10, 100)
x2 = -w[0]/w[2] - (w[1]/w[2])*x1
plt.plot(x1,x2,'k')
plt.scatter(X['SEPAL_LENGHT'],X['PETAL_LENGHT'])
plt.show()
def simple_example():
df = pd.read_csv(
'https://archive.ics.uci.edu/ml/machine-learning-databases/iris/iris.data',
header=None)
dummy_data = df[48:52]
y = dummy_data.iloc[:, 4].values
y = np.where(y == 'Iris-setosa', -1, 1)
X = dummy_data.iloc[:, [0, 2]].values
X[0, 1] = 3
plt.scatter(X[:2, 0], X[:2, 1], color='red', marker='o', label='setosa')
plt.scatter(X[2:, 0],
X[2:, 1],
color='blue',
marker='x',
label='versicolor')
plt.xlabel('sepal length')
plt.ylabel('petal length')
plt.legend(loc='upper left')
plt.show()
ppn = Perceptron(eta=0.01, n_iter=40)
ppn.fit(X, y)
plt.plot(range(1, len(ppn.errors_) + 1), ppn.errors_, marker='o')
plt.xlabel('Epochs')
plt.ylabel('Number of misclassifications')
plt.show()
plot_decision_regions(X, y, classifier=ppn)
plt.xlabel('sepal length [cm]')
plt.ylabel('petal length [cm]')
plt.legend(loc='upper left')
plt.show()
def perceptron_model(learning_rate=0.01, n_iters=1000, split_test_ratio=0.2):
X, y = datasets.make_blobs(n_samples=150,
n_features=2,
centers=2,
cluster_std=1.05,
random_state=2)
X_train, X_test, y_train, y_test = train_test_split(
X, y, test_size=split_test_ratio, random_state=123)
p = Perceptron(learning_rate=0.01, n_iters=1000)
p.fit(X_train, y_train)
predictions = p.predict(X_test)
acc = accuracy(y_test, predictions)
print("Perceptron classification accuracy", acc)
fig = plt.figure()
ax = fig.add_subplot(1, 1, 1)
plt.scatter(X_train[:, 0], X_train[:, 1], marker='o', c=y_train)
x0_1 = np.amin(X_train[:, 0])
x0_2 = np.amax(X_train[:, 0])
x1_1 = (-p.weights[0] * x0_1 - p.bias) / p.weights[1]
x1_2 = (-p.weights[0] * x0_2 - p.bias) / p.weights[1]
ax.plot([x0_1, x0_2], [x1_1, x1_2], 'k')
ymin = np.amin(X_train[:, 1])
ymax = np.amax(X_train[:, 1])
ax.set_ylim([ymin - 3, ymax + 3])
plt.show()
return predictions, acc, plt
def test_score(self):
expected_score = 1.0
design_matrix, target_values = self.get_design_matrix_and_target_values('OR')
perceptron = Perceptron(max_iter=100, learning_rate=0.2, activation_function='heaviside', seed=0)
perceptron.fit(design_matrix, target_values)
score = perceptron.score(design_matrix, target_values)
self.assertEqual(expected_score, score)
def run_train_validation_perceptron(rate, epochs):
perceptron4 = Perceptron(rate, epochs, "Perceptron 4 Train & Validation")
perceptron4.fit(train_and_valid_images, train_and_valid_labels)
final_results = perceptron4.predict(test_images)
perceptron4.calculate_results(final_results, test_labels, 150)
perceptron4.graph_perceptron()
def Q4():
for m in m_num:
class_check = True
X = None
y = None
while class_check:
X = np.random.multivariate_normal(mean, cov, m)
y = np.sign(np.array([0.3, -0.5]) @ X.T + 0.1)
if np.abs(np.sum(y)) == m:
class_check = True
else:
class_check = False
plt.scatter(X.T[0], X.T[1], c=y)
pts = np.linspace(-3.5, 3.5, 1000)
plt.plot(pts, 0.6 * pts + 0.2, label='True')
perceptron = Perceptron()
perceptron.fit(X, y)
w_perceptron = perceptron.w
plt.plot(pts,
-(w_perceptron[0] * pts + w_perceptron[2]) / w_perceptron[1],
label='Perceptron')
clf = SVC(C=1e10, kernel='linear')
clf.fit(X, y)
w_clf = clf.coef_[0]
plt.plot(pts,
-(w_clf[0] * pts + clf.intercept_[0]) / w_clf[1],
label='SVM')
plt.legend()
plt.title('Hyperplane Classification for %i Samples' % m)
plt.show()
def ch2():
# df = pd.read_csv("https://archive.ics.uci.edu/ml/machine-learning-databases/iris/iris.data", header=None)
df = pd.read_csv(
"http://mlr.cs.umass.edu/ml/machine-learning-databases/iris/iris.data",
header=None)
y = df.iloc[0:100, 4].values
y = np.where(y == 'Iris-setosa', -1, 1)
X = df.iloc[0:100, [0, 2]].values
plt.scatter(X[:50, 0], X[:50, 1], color='red', marker='o', label='setosa')
plt.scatter(X[50:100, 0],
X[50:100, 1],
color='blue',
marker='x',
label='versicolor')
ppn = Perceptron(eta=0.1, n_iter=10)
ppn.fit(X, y)
plt.xlabel('sepal length [cm]')
plt.ylabel('petal length [cm]')
plt.legend(loc='upper left')
plot_decision_regions(X, y, classifier=ppn)
plt.savefig(PIC_LOC + "iris_ch2.png")
plt.close()
plt.plot(range(1, len(ppn.errors_) + 1), ppn.errors_, marker='o')
plt.xlabel('Epochs')
plt.ylabel('Number of misclassifications')
plt.savefig(PIC_LOC + "iris2_ch2.png")
plt.close()
def main():
X, Y = loadIris()
p = Perceptron(learnRate=0.1, learnAccuracy=1.0)
p.fit(X, Y)
plotDecisionBoundary(p, X, Y)
print(p.predict([5.0, 2.1]))
def train_perceptron(train_X, train_y, saved_model_path):
train_size = train_X.shape[0]
feat_dim = train_X.shape[1]
epoch_num = 15
lr = 0.0001
clf = Perceptron(input_dim=feat_dim, lr=lr, epoch_num=epoch_num)
clf.fit(train_X, train_y)
clf.save_model(saved_model_path)
def iris(plotting_results, live_plotting):
# import and ready input file
input_file = "iris.csv"
df = pd.read_csv(input_file, header=None)
df.head()
# X: values, y: targets
# extract features
X = df.iloc[:, 0:4].values
# extract the label column
y = df.iloc[:, 4].values
# Setosa:
y_setosa = np.where(y == 'Setosa', 1, 0)
# Versicolor:
y_versicolor = np.where(y == 'Versicolor', 1, 0)
# Virginica:
y_virginica = np.where(y == 'Virginica', 1, 0)
sets = [y_setosa, y_versicolor, y_virginica]
predictions = []
y_test = []
i = 0
for set in sets:
# split data into train and test sets
X_train, X_test, y_train, y_test_tmp = train_test_split(
X, set, test_size=0.2, random_state=123)
y_test.append(y_test_tmp)
# create and train model
p = Perceptron(learning_rate=0.01, n_iters=300)
p.fit(X_train, y_train, plot=live_plotting)
predictions.append(p.predict(X_test))
# predictions: exodos, y_test: stoxos
print("Perceptron classification accuracy",
accuracy(y_test[i], predictions[i]) * 100, "%")
i += 1
if (plotting_results):
fig, (ax) = plt.subplots(1, 3, sharex=True, sharey=True)
fig.suptitle("Results")
for i in range(3):
ax[i].scatter(
range(len(y_test[i])), y_test[i], marker='o',
color='b') # mple teleies: pragmatikoi stoxoi (y_test)
ax[i].scatter(range(len(predictions[i])),
predictions[i],
marker='.',
color='r') # kokkinoi kykloi: exwdos (predictions)
ax[i].set_xlabel("protypo")
ax[i].set_ylabel("exodos (r) / stoxos (b)")
def __test_perceptron(self, perceptron: Perceptron):
samples = numpy.array([[3, 3], [4, 3], [1, 1]])
labels = numpy.array([1, 1, -1])
perceptron.fit(samples, labels)
for i in range(samples.shape[0]):
self.assertEqual(perceptron.predict(samples[i]), labels[i])
logging.debug("i = {} success".format(i))
def step2_learning():
ppn = Perceptron(eta=0.1)
X, y = step1_get_data()
ppn.fit(X, y)
print(ppn.errors_)
print(ppn.w_)
with open('perceptron.dat', 'wb') as fp:
pickle.dump(ppn, fp)
print('학습 완료')
def accuracy(y_true, y_pred):
accuracy = np.sum(y_true == y_pred) / len(y_true)
return accuracy
X, y = datasets.make_blobs(n_samples=150,n_features=2,centers=2,cluster_std=1.05,random_state=2)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=123)
p = Perceptron(learning_rate=0.01, n_iters=1000)
p.fit(X_train, y_train)
predictions = p.predict(X_test)
def q2(X, y):
print('Solving q2')
model = Perceptron()
model.fit(X,
y,
alpha=0.001,
weight_init='random',
epochs=200,
verbose=False,
do_plot=True)
def main():
# Training data for logical OR function
training_data = np.array([[0, 0, 0], [0, 1, 1], [1, 0, 1], [1, 1, 1]])
design_matrix = training_data[:, :2]
target_values = training_data[:, -1]
perceptron = Perceptron(max_iter=100, learning_rate=0.2)
perceptron.fit(design_matrix, target_values)
predictions = perceptron.predict(design_matrix)
print(predictions)
def main():
plotting = int(input("0. Live Plot\n1. Plot Results\n2. Both\n") or 2)
live_plotting = plotting == 0 or plotting == 2
plotting_results = plotting == 1 or plotting == 2
live_plotting_3d, plotting_results_3d = False, False
file = input(
"Δώσε input file (a, b, c, d, ii_a, ii_b, iris, bitmap): ") or 'a'
if (file == "iris"):
iris(plotting_results, live_plotting)
elif (file == "bitmap"):
bitmap(plotting_results, live_plotting)
else:
input_file = 'data_package_%s.csv' % file
if (file.__contains__("ii_")):
live_plotting_3d, plotting_results_3d = live_plotting, plotting_results
live_plotting, plotting_results = False, False
df = pd.read_csv(input_file, header=0)
df = df._get_numeric_data()
# targets
targets_file = 'data_package_values_%s.csv' % file
targets_df = pd.read_csv(targets_file, header=0)
targets_df = targets_df._get_numeric_data()
# x: values, y: targets
X = df.values
y = targets_df.values
# split data into train and test sets
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.2,
random_state=123)
# create and train model
p = Perceptron(learning_rate=0.01, n_iters=100)
p.fit(X_train, y_train, plot=live_plotting, plot_3d=live_plotting_3d)
predictions = p.predict(X_test)
if (plotting_results):
fig = plt.figure()
ax = fig.add_subplot(1, 1, 1)
# mple teleies: pragmatikoi stoxoi (y_test)
plt.scatter(range(len(y_test)), y_test, marker='o', color='b')
plt.scatter(range(len(predictions)),
predictions,
marker='.',
color='r') # kokkinoi kykloi: exwdos (predictions)
plt.xlabel("protypo")
plt.ylabel("exodos (r) / stoxos (b)")
if (plotting_results_3d):
plot_results_3d(p, X_test, y_test, predictions)
print("Perceptron classification accuracy",
accuracy(y_test, predictions), "%")
plt.show()
def main():
x = [(-2.8, 1.4), (-0.2, -3.5), (2.8, -4), (-2.1, -2.7), (0.3, -4.1),
(-1, -4)]
y = [0, 1, 1, 0, 1, 0]
class_a_x = []
class_a_y = []
class_b_x = []
class_b_y = []
for i, p in enumerate(x):
if y[i] == 0:
class_a_x.append(p[0])
class_a_y.append(p[1])
else:
class_b_x.append(p[0])
class_b_y.append(p[1])
print(class_a_x)
print(class_a_y)
colors_a = (1, 0, 0)
colors_b = (0, 0, 1)
area = m.pi * 3
plt.scatter(class_a_x, class_a_y, s=area, color=colors_a, alpha=0.5)
plt.scatter(class_b_x, class_b_y, s=area, color=colors_b, alpha=0.5)
plt.title('Scatter plot')
plt.xlabel('x')
plt.ylabel('y')
classifier = Perceptron()
classifier.fit(np.array(x), np.array(y), n_epochs=200)
for p, ref in zip(x, y):
if (classifier.predict(p) != ref):
print(f'loser! WRONG CLASSIFICATION: {p}')
break
weights = classifier.get_weights()
k = -weights[0] / weights[1]
b = -weights[2] / weights[1]
print(f'{k=}')
print(f'{b=}')
f = lambda x_: k * x_ + b
x_graph = np.linspace(-5, 5, 2)
y_graph = np.array(list(map(f, x_graph)))
plt.plot(x_graph, y_graph, 'g-', linewidth=2, markersize=12)
plt.show()
def main():
# load data into shuffle matrix
df = pd.read_csv("diabetes.csv")
data = np.array(df)
np.random.shuffle(data)
# split data into labels and design matrix
design = data[:, :-1].T
labels = data[:, -1]
labels = labels.reshape((labels.shape[0],1)).T # keep dims to 2
n, m_tot = design.shape
# normalize dataset
maxs = np.amax(design, axis=1)
mins = np.amin(design, axis=1)
ranges = maxs - mins
design -= mins.reshape(len(mins),1)
design /= ranges.reshape(len(ranges),1)
# split into test and training data
frac_test = .8
split_idx = int(frac_test*m_tot)
train_design = design[:, :split_idx]
train_labels = labels[:, :split_idx]
test_design = design[:, split_idx:]
test_labels = labels[:, split_idx:]
# fit neural network
nn = NN(ns=[n,5,1], acts=["ReLU","ReLU","sigmoid"])
nn.fit(train_design, train_labels, alpha=1e-2, epochs = 20_000)
test_acc = nn.evaluate(X=test_design, Y=test_labels)
train_acc = nn.evaluate(X=train_design, Y=train_labels)
print("Network test set accuracy: %.5f" % test_acc)
print("Network training set accuracy: %.5f" % train_acc)
print()
# fit perceptron
perc = Perceptron()
perc.fit(X=train_design, Y=train_labels, alpha=1e-4,
lambd=1e-2, epochs=100_000)
test_acc = perc.acc(X=test_design, Y=test_labels)
train_acc = perc.acc(X=train_design, Y=train_labels)
print("Own perceptron test set accuracy: %.5f" % test_acc)
print("Own perceptron training set accuracy: %.5f" % train_acc)
print()
# fit standard template perceptron from sklearn
clf = standardPerceptron(tol=1e-3, random_state=0)
clf.fit(train_design.T, train_labels.squeeze())
train_acc = clf.score(train_design.T, train_labels.squeeze())
test_acc = clf.score(test_design.T, test_labels.squeeze())
print("Sklearn perceptron test set accuracy: %.5f" % test_acc)
print("Sklearn perceptron training set accuracy: %.5f"%train_acc)
print()
def one_vs_five():
X, y = get_digits()
X_train, X_test, y_train, y_test = train_test_split(X,
y,
train_size=0.8,
shuffle=False)
c = Perceptron(iterations=100)
c.fit(X_train, y_train)
pred = c.predict(X_train).reshape(-1)
visualize(X_train, y_train, pred, c.w.reshape(-1), DIGITS_PLOT)
print("Accuracy:", np.mean(pred == y_test))
def test_e22(self):
logger.info("test case e22")
# data e2.1
data_raw = np.loadtxt("Input/data_2-1.txt")
X = data_raw[:, :2]
y = data_raw[:, -1]
clf = Perceptron(verbose=False)
clf.fit(X, y)
y_pred = clf.predict(X)
logger.info(clf.w)
logger.info(str(y_pred))
self.assertListEqual(y.tolist(), y_pred.tolist())
def train(X, y):
classifier = Perceptron(learning_rate=0.1, n_epochs=10)
classifier.fit(X, y)
plt.figure(2)
plt.title('Prediction error decreasing over time')
time_steps = range(1, len(classifier.get_errors()) + 1)
plt.plot(time_steps, classifier.get_errors(), marker='o')
plt.xlabel('Epochs')
plt.ylabel('Number of weight updates'
) # we only update when there is an error in prediction
print('Number of errors per epoch (should be decreasing)')
plot_decision_regions(X, y, classifier=classifier)
plt.show()
def run_pca50_perceptron():
# Try PCA with 50 features
pca2 = PCA(n_components=50)
pca2.fit(train_images)
pca_train_images2 = pca2.transform(train_images)
pca_test_images2 = pca2.transform(test_images)
# try perceptron with 50 dimensions
perceptron3 = Perceptron(0.01, 15, "Perceptron 3 PCA - 50")
perceptron3.fit(pca_train_images2, train_labels)
perceptron3.graph_perceptron()
results_with_pca2 = perceptron3.predict(pca_test_images2)
perceptron3.calculate_results(results_with_pca2, test_labels, 150)
def my_model(X_train, y_train, X_test, y_test):
from perceptron import Perceptron
constant = np.ones((len(X_train), 1))
X_train = np.hstack((constant, X_train))
const = np.ones((len(X_test), 1))
X_test = np.hstack((const, X_test))
lr = Perceptron(0.0004, 1000)
lr.fit(X_train, y_train)
y_pred_test = lr.predict(X_test)
acc_test = accuracy_score(y_test, y_pred_test)
prec_test = precision_score(y_test, y_pred_test, average='micro')
recall_test = recall_score(y_test, y_pred_test, average='micro')
return acc_test, prec_test, recall_test
def step2_learning():
ppn = Perceptron(eta=0.1)
data = step1_get_data()
X = data[0]
y = data[1]
# 학습한다.
ppn.fit(X, y)
print(ppn.errors_)
print(ppn.w_)
# 학습된 객체를지정한다.
# 학습이 완료된 객체를 파일로 저장한다.
with open('./3.IrisPerceptron/perceptron.dat', 'wb') as fp:
pickle.dump(ppn, fp)
print("학습 완료")
def main():
data = DataManager.load_data("data/data_banknote_authentication.txt")
# remove break lines
data = np.array([item.strip("\n") for item in data])
data = np.array([item.split(',') for item in data])
data = data.astype(np.float)
# 0 for authentic and 1 for inauthentic
df = pd.DataFrame(data)
df[4] = df[4].astype(int)
authentic = df[df[4] == 0]
inauthentic = df[df[4] == 1]
X = df.iloc[np.r_[0:200, 1100:1300], [0, 3]].values
y = [0 if x < 200 else 1 for x in range(400)]
plt.scatter(X[:200, 0],
X[:200, 1],
color='red',
marker='o',
label='authentic')
plt.scatter(X[200:400, 0],
X[200:400, 1],
color='blue',
marker='x',
label='inauthentic')
plt.xlabel('variance of Wavelet Transformed image')
plt.ylabel('entropy of image')
plt.legend(loc='upper left')
plt.show()
ppn = Perceptron(eta=0.1, n_iter=10)
ppn.fit(X, y)
plot_decision_regions(X, y, clasiifier=ppn)
plt.xlabel('variance of Wavelet Transformed image')
plt.ylabel('entropy of image')
plt.legend(loc='upper left')
plt.show()
X_std = np.copy(X)
X_std[:, 0] = (X[:, 0] - X[:, 0].mean()) / X[:, 0].std()
X_std[:, 1] = (X[:, 1] - X[:, 1].mean()) / X[:, 1].std()
ada = AdalineGD(n_iter=10, eta=0.01)
ada.fit(X_std, y)
plot_decision_regions(X_std, y, clasiifier=ada)
plt.title('Adaline - gradient descent')
plt.xlabel('variance of Wavelet Transformed image')
plt.ylabel('entropy of image')
plt.legend(loc='upper left')
plt.show()
def perceptron_model():
""" Perceptron classifier on Iris flower dataset
"""
df = pd.read_csv('https://archive.ics.uci.edu/ml/machine-learning-databases/iris/iris.data', header=None)
# setosa and versicolor
y = df.iloc[0:100, 4].values
y = np.where(y == 'Iris-setosa', -1, 1)
# sepal length and petal length
X = df.iloc[0:100, [0, 2]].values
ppn = Perceptron(epochs=10, eta=0.1)
ppn.fit(X, y)
print('Weights: %s' % ppn.w_)
def step2_learing():
ppn = Perceptron(eta=0.1)
# print(ppn)
data = step1_get_data()
X = data[0] #꽃잎 길이와 너비
y = data[1] #품종
# 학습한다
ppn.fit(X, y)
print(ppn.errors_)
print(ppn.w_)
# 학습된 객체를 저장한다
# 학습이 완료된 객체를 파일로 저장한다.
with open('./perceptron.dat', 'wb') as fp:
pickle.dump(ppn, fp)
print('학습완료')
def train_perceptron(X, y):
"""Training the perceptron model"""
ppn = Perceptron(eta=0.1, n_iter=10)
ppn.fit(X, y)
plt.plot(range(1, len(ppn.errors_) + 1), ppn.errors_, marker='o')
plt.xlabel('Epochs')
plt.ylabel('Number of updates')
plt.tight_layout()
# plt.savefig('./perceptron_1.png', dpi=300)
plt.show()
plot_decision_regions(X, y, classifier=ppn)
plt.xlabel('sepal length [cm]')
plt.ylabel('petal length [cm]')
plt.legend(loc='upper left')
plt.tight_layout()
# plt.savefig('./perceptron_2.png', dpi=300)
plt.show()
support_vector_machine = SupportVectorMachine(C=1, kernel=rbf_kernel)
# ........
# TRAIN
# ........
print "Training:"
print "\tAdaboost"
adaboost.fit(X_train, rescaled_y_train)
print "\tNaive Bayes"
naive_bayes.fit(X_train, y_train)
print "\tLogistic Regression"
logistic_regression.fit(X_train, y_train)
print "\tMultilayer Perceptron"
mlp.fit(X_train, y_train, n_iterations=20000, learning_rate=0.1)
print "\tPerceptron"
perceptron.fit(X_train, y_train)
print "\tDecision Tree"
decision_tree.fit(X_train, y_train)
print "\tRandom Forest"
random_forest.fit(X_train, y_train)
print "\tSupport Vector Machine"
support_vector_machine.fit(X_train, rescaled_y_train)
# .........
# PREDICT
# .........
y_pred = {}
y_pred["Adaboost"] = adaboost.predict(X_test)
y_pred["Naive Bayes"] = naive_bayes.predict(X_test)
y_pred["K Nearest Neighbors"] = knn.predict(X_test, X_train, y_train)
y_pred["Logistic Regression"] = logistic_regression.predict(X_test)
# color='red', marker='o', label='setosa')
# # 品種 versicolorのプロット
# plt.scatter(X[50:100, 0], X[50:100, 1],
# color='blue', marker='x', label='versicolor')
# # 軸ラベルの設定
# plt.xlabel('sepal length [cm]')
# plt.ylabel('petal length [cm]')
# # 凡例の設定
# plt.legend(loc='upper left')
# # 図の表示
# plt.show()
# パーセプトロンのオブジェクト生成
ppn = Perceptron(eta=0.1, n_iter=10)
# トレーニングデータへのモデルの適合
ppn.fit(X, y)
# エポックと誤分類誤差の関係の折れ線グラフをプロット
# plt.plot(range(1, len(ppn.errors_) + 1), ppn.errors_, marker='o')
# # 軸のラベルの設定
# plt.xlabel('Epochs')
# plt.ylabel('Number of misclassifications')
# # 図の表示
# plt.show()
from matplotlib.colors import ListedColormap
def plot_decision_regions(X, y, classifier, resolution=0.02):
# マーカーとカラーマップの準備
markers = ('s', 'x', 'o', '^', 'v')
colors = ('red', 'blue', 'lightgreen', 'gray', 'cyan')
ax = fig1.add_subplot(1, 1, 1)
# setosa iris
ax.scatter(X["Sepal length"].iloc[:50], X["Petal length"].iloc[:50],
color="red", marker="o", label="Iris setosa")
# versicolor iris
ax.scatter(X["Sepal length"].iloc[50:], X["Petal length"].iloc[50:],
color="blue", marker="o", label="Iris versicolor")
ax.set(xlabel="Sepal lenght", ylabel="Petal length")
ax.legend(loc="upper left")
# train the perceptron classifier
iris_ppn = Perceptron()
iris_ppn.fit(X.values, y.values)
# plot the numbers of errors for each perceptron iteration
fig2 = plt.figure()
ax = fig2.add_subplot(1, 1, 1)
ax.plot(range(1, iris_ppn.n_iterations + 1), iris_ppn.errors_, marker="o")
ax.set(xlabel="Iterations", ylabel="Number of missclassifications")
# plot the decision boundaries for the 2D dataset
fig3, ax = plot_decision_regions(X.values, y.values, iris_ppn)
ax.set(xlabel="sepal length [cm]", ylabel= "petal length [cm]")
ax.legend(loc="upper left")
#!/usr/bin/python #-*-coding:utf-8-* from perceptron import Perceptron from gen_data import gen_arti, plot_frontiere, plot_data import matplotlib.pyplot as plt ### Generer et tracer des donnees DATAX, DATAY = gen_arti(data_type=0, nbex=1000, eps=0.1) PERCEP = Perceptron(eps=1e-1, max_iter=1000) PERCEP.fit(DATAX, DATAY) print PERCEP.score(DATAX, DATAY) plot_frontiere(DATAX, PERCEP.predict, 50) plot_data(DATAX, DATAY) plt.show()
# plot data
plt.scatter(training_features[:50, 0], training_features[:50, 1],
color='red', marker='o', label='setosa')
plt.scatter(training_features[50:100, 0], training_features[50:100, 1],
color='blue', marker='x', label='versicolor')
plt.xlabel('petal length [cm]')
plt.ylabel('sepal length [cm]')
plt.legend(loc='upper left')
plt.tight_layout()
# plt.savefig('./iris_1.png', dpi=300)
plt.show()
ppn = Perceptron(eta=0.1, n_iter=10)
ppn.fit(training_features, targets)
plt.plot(range(1, len(ppn.errors_) + 1), ppn.errors_, marker='o')
plt.xlabel('Epochs')
plt.ylabel('Number of misclassifications')
plt.tight_layout()
# plt.savefig('./perceptron_1.png', dpi=300)
plt.show()
plot_decision_regions(training_features, targets, classifier=ppn)
plt.xlabel('sepal length [cm]')
plt.ylabel('petal length [cm]')
plt.legend(loc='upper left')
plt.tight_layout()