def plot_ridge_n_samples():
X, y = load_extended_boston()
plot_learning_curve(Ridge(alpha=1), X, y)
plot_learning_curve(LinearRegression(), X, y)
plt.legend(loc=(0, 1.05), ncol=2, fontsize=11)
plt.show()
def linear_regression_boston():
X, y = load_extended_boston()
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)
ridge = Ridge().fit(X_train, y_train)
print("w[0]: %f b: %f" % (ridge.coef_[0], ridge.intercept_))
print("Training set score : {:.2f}".format(ridge.score(X_train, y_train)))
print("Test set score ] {:.2f}".format(ridge.score(X_test, y_test)))
def linear_regression_boston(alpha):
X, y = load_extended_boston()
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)
lasso = Lasso(alpha=alpha, max_iter=100000).fit(X_train, y_train)
print("-------------------")
print("[alpha : {}]Training set score: {:.2f}".format(alpha, lasso.score(X_train, y_train)))
print("[alpha : {}]Test set score: {:.2f}".format(alpha, lasso.score(X_test, y_test)))
print("[alpha : {}]Number of features used: {}".format(alpha, np.sum(lasso.coef_ != 0)))
def linear_regression_housing():
X, y = load_extended_boston()
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)
line = np.linspace(-3, 3, 100).reshape(-1, 1)
lr = LinearRegression().fit(X_train, y_train)
print("w[0]: %f b: %f" % (lr.coef_[0], lr.intercept_))
print("Training set score : {:.2f}". format(lr.score(X_train, y_train)))
print("Test set score ] {:.2f}".format(lr.score(X_test, y_test)))
def linear_regression_boston():
X, y = load_extended_boston()
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)
limit = np.arange(0.01, 0.50, 0.01)
training_scores = np.zeros(len(limit), dtype=float)
test_scores = np.zeros(len(limit), dtype=float)
for i, alpha in enumerate(limit):
training_scores[i], test_scores[i] = ridge_fit(alpha, X_train, X_test,
y_train, y_test)
plt.plot(limit, test_scores, label="Test Score")
plt.plot(limit, training_scores, label="Tranining Score")
plt.legend()
plt.show()
def linear_regression_boston():
X, y = load_extended_boston()
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)
limit = np.arange(0.0001, 1.0001, 0.001)
training_scores = np.zeros(len(limit), dtype=float)
test_scores = np.zeros(len(limit), dtype=float)
best_score = 0.0
best_alpha = 0.0
for i, alpha in enumerate(limit):
training_scores[i], test_scores[i] = lasso_fit(alpha, X_train, X_test,
y_train, y_test)
if best_score < test_scores[i]:
best_score = test_scores[i]
best_alpha = alpha
print("Best Alpha : {}, Best Score : {}".format(best_alpha, best_score))
plt.plot(limit, test_scores, label="Test Score")
plt.plot(limit, training_scores, label="Tranining Score")
plt.legend()
plt.show()
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Thu May 30 11:22:00 2019
@author: prawigya
"""
from mglearn.datasets import load_extended_boston
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split
import matplotlib.pyplot as plt
import numpy as np
X, y = load_extended_boston()
x_train, x_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)
lr = LinearRegression().fit(x_train, y_train)
print("Accuracy of train data: ", lr.score(x_train, y_train))
print("Accuracy of test data: ", lr.score(x_test, y_test
lasso1 = Lasso(alpha = 0.1).fit(x_train, y_train)
print("Training score ", lasso1.score(x_train, y_train))
print("Test score ", lasso1.score(x_test, y_tetst))
def r_square(y_true, y_pred):
from keras import backend as K
SS_res = K.sum(K.square(y_true - y_pred))
SS_tot = K.sum(K.square(y_true - K.mean(y_true)))
return (1 - SS_res/(SS_tot + K.epsilon()))
# R^2로 설명할 수 없는 부분
def r_square_loss(y_true, y_pred):
from keras import backend as K
SS_res = K.sum(K.square(y_true - y_pred))
SS_tot = K.sum(K.square(y_true - K.mean(y_true)))
return 1 - ( 1 - SS_res/(SS_tot + K.epsilon()))
# 1) data loading
from mglearn.datasets import load_extended_boston
df_boston_et_x, df_boston_et_y = load_extended_boston()
# 2) scaling
m_sc = standard()
m_sc.fit(df_boston_et_x)
df_boston_et_x = m_sc.transform(df_boston_et_x)
# 3) data split
train_x, test_x, train_y, test_y = train_test_split(df_boston_et_x,
df_boston_et_y,
random_state=0)
# 4) 모델 생성
nx = df_boston_et_x.shape[1]
model = Sequential()
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Tue Dec 18 01:11:10 2018
@author: Nihat Allahverdiyev
"""
import mglearn.datasets as md
from sklearn.linear_model import Ridge
from sklearn.model_selection import train_test_split
import matplotlib.pyplot as plt
from sklearn.linear_model import LinearRegression
X, y = md.load_extended_boston()
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state= 0)
lr = LinearRegression().fit(X_train, y_train)
ridge = Ridge().fit(X_train, y_train)
print("Accuracy of predicting train data: {:.2f}".format(ridge.score(
X_train, y_train)))
print("Accuracy of predicting test data: {:.2f}".format(ridge.score(
X_test, y_test)))
ridge10 = Ridge(alpha = 10).fit(X_train, y_train)
ridge01 = Ridge(alpha = 0.1).fit(X_train, y_train)
plt.plot(ridge.coef_, 's', label = "Ridge alpha = 1")
from mglearn.datasets import load_extended_boston
from sklearn.model_selection import train_test_split
from sklearn.linear_model import Lasso, Ridge
import numpy as np
boston_data, boston_target = load_extended_boston()
x_train, x_test, y_train, y_test = train_test_split(boston_data,
boston_target,
random_state=0,
test_size=0.3)
lasso = Lasso().fit(x_train, y_train)
print('{:.3f}'.format(lasso.score(x_train, y_train)))
# 0.265
print('{:.3f}'.format(lasso.score(x_test, y_test)))
# 0.214
print(lasso.coef_)
# array([-0. , 0. , -0. , 0. , -0. ,
# 0. , -0. , 0. , -0. , -0. ,
from mglearn import datasets
from sklearn.datasets import load_boston
boston = load_boston()
print("데이터의 형태: {}".format(boston.data.shape))
X, y = datasets.load_extended_boston()
print("X.shape: {}".format(X.shape))
