sklearn 简介
机器学习 Machine Learning
监督学习 supervised learning;
非监督学习 unsupervised learning;
半监督学习 semi-supervised learning;
强化学习 reinforcement learning;
遗传算法 genetic algorithm.
sklearn 安装
Anaconda安装
pip安装
sklearn 一般使用
选择学习方法
通用学习方式
sklearn 强大数据库
sklearn 常用属性与功能
选择学习方法
看图选方法
常用算法可分为四类：分类，回归，聚类，降维。
img1
通用学习方式
要点
创建数据
建立模型－训练－预测
要点：用KNN classifier，对Iris数据集进行分类
创建数据：
from sklearn import datasetsfrom sklearn.model_selection import train_test_splitfrom sklearn.neighbors import KNeighborsClassifier iris = datasets.load_iris() iris_X = iris.data iris_y = iris.target print(iris_X[:2, :]) print(iris_y)""" [[ 5.1 3.5 1.4 0.2] [ 4.9 3. 1.4 0.2]] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2] """ X_train, X_test, y_train, y_test = train_test_split( iris_X, iris_y, test_size=0.3) print(y_train)""" [2 1 0 1 0 0 1 1 1 1 0 0 1 2 1 1 1 0 2 2 1 1 1 1 0 2 2 0 2 2 2 2 2 0 1 2 2 2 2 2 2 0 1 2 2 1 1 1 0 0 1 2 0 1 0 1 0 1 2 2 0 1 2 2 2 1 1 1 1 2 2 2 1 0 1 1 0 0 0 2 0 1 0 0 1 2 0 2 2 0 0 2 2 2 1 2 0 0 2 1 2 0 0 1 2] """
建立模型－训练－预测:
knn = KNeighborsClassifier() knn.fit(X_train, y_train) print(knn.predict(X_test)) print(y_test)""" [2 0 0 1 2 2 0 0 0 1 2 2 1 1 2 1 2 1 0 0 0 2 1 2 0 0 0 0 1 0 2 0 0 2 1 0 1 0 0 1 0 1 2 0 1] [2 0 0 1 2 1 0 0 0 1 2 2 1 1 2 1 2 1 0 0 0 2 1 2 0 0 0 0 1 0 2 0 0 2 1 0 1 0 0 1 0 1 2 0 1] """
sklearn 强大数据库
要点
导入数据－训练模型
创建虚拟数据－可视化
要点：使用 sklearn 读取数据库和生成虚拟的数据，例如用来训练线性回归模型的数据
img1
img2
sklearn.datasets.make_regression(n_samples=100, n_features=100, n_informative=10, n_targets=1, bias=0.0, effective_rank=None, tail_strength=0.5, noise=0.0, shuffle=True, coef=False, random_state=None)[source]
导入数据－训练模型：
from __future__ import print_functionfrom sklearn import datasetsfrom sklearn.linear_model import LinearRegressionimport matplotlib.pyplot as plt loaded_data = datasets.load_boston() data_X = loaded_data.data data_y = loaded_data.target model = LinearRegression() model.fit(data_X, data_y) print(model.predict(data_X[:4, :])) print(data_y[:4]) “”“ [ 30.00821269 25.0298606 30.5702317 28.60814055] [ 24. 21.6 34.7 33.4] ”“”
创建虚拟数据－可视化:
X, y = datasets.make_regression(n_samples=100, n_features=1, n_targets=1, noise=10) plt.scatter(X, y) plt.show()
img3
X, y = datasets.make_regression(n_samples=100, n_features=1, n_targets=1, noise=50) plt.scatter(X, y) plt.show()
img4
sklearn 常用属性与功能
model.coef_:输出模型的斜率model.intercept_:输出模型的截距（与y轴的交点）model.get_params():取出之前定义的参数model.score(data_X, data_y):对 Model 用 R^2 的方式进行打分
训练和预测:
from sklearn import datasetsfrom sklearn.linear_model import LinearRegression loaded_data = datasets.load_boston() data_X = loaded_data.data data_y = loaded_data.target model = LinearRegression() model.fit(data_X, data_y) print(model.predict(data_X[:4, :]))""" [ 30.00821269 25.0298606 30.5702317 28.60814055] """
参数和分数:
print(model.coef_)
print(model.intercept_)"""
[ -1.07170557e-01 4.63952195e-02 2.08602395e-02 2.68856140e+00
-1.77957587e+01 3.80475246e+00 7.51061703e-04 -1.47575880e+00
3.05655038e-01 -1.23293463e-02 -9.53463555e-01 9.39251272e-03
-5.25466633e-01]
36.4911032804
"""print(model.get_params())"""
{'copy_X': True, 'normalize': False, 'n_jobs': 1, 'fit_intercept': True}
"""print(model.score(data_X, data_y)) # R^2 coefficient of determination"""
0.740607742865
"""sklearn 高级使用
正规化 Normalization
检查神经网络（Evaluation）
交叉验证 1 Cross-validation
交叉验证 2 Cross-validation
交叉验证 3 Cross-validation
保存模型
正规化 Normalization
数据标准化
数据标准化对机器学习成效的影响
数据标准化:
from sklearn import preprocessing #标准化数据模块import numpy as np#建立Arraya = np.array([[10, 2.7, 3.6], [-100, 5, -2], [120, 20, 40]], dtype=np.float64)#将normalized后的a打印出print(preprocessing.scale(a))# [[ 0. -0.85170713 -0.55138018]# [-1.22474487 -0.55187146 -0.852133 ]# [ 1.22474487 1.40357859 1.40351318]]
数据标准化对机器学习成效的影响:
# 标准化数据模块from sklearn import preprocessing import numpy as np# 将资料分割成train与test的模块from sklearn.model_selection import train_test_split# 生成适合做classification资料的模块from sklearn.datasets.samples_generator import make_classification # Support Vector Machine中的Support Vector Classifierfrom sklearn.svm import SVC # 可视化数据的模块import matplotlib.pyplot as plt #生成具有2种属性的300笔数据X, y = make_classification( n_samples=300, n_features=2, n_redundant=0, n_informative=2, random_state=22, n_clusters_per_class=1, scale=100)#可视化数据plt.scatter(X[:, 0], X[:, 1], c=y) plt.show()
img1
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3) clf = SVC() clf.fit(X_train, y_train)print(clf.score(X_test, y_test))# 0.477777777778
img1
X = preprocessing.scale(X) X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3) clf = SVC() clf.fit(X_train, y_train)print(clf.score(X_test, y_test))# 0.9
检查神经网络（Evaluation）
Training and Test data
误差曲线
准确度曲线
正规化
交叉验证
Training and Test data：70% training,30% testing
img1
误差曲线:
img2
准确度曲线:
img3
正规化:
img4
交叉验证:
img5
交叉验证 1 Cross-validation
Model 基础验证法
Model 交叉验证法(Cross Validation)
以准确率(accuracy)判断
以平均方差(Mean squared error)
Model 基础验证法:
from sklearn.datasets import load_iris # iris数据集from sklearn.model_selection import train_test_split # 分割数据模块from sklearn.neighbors import KNeighborsClassifier # K最近邻(kNN，k-NearestNeighbor)分类算法#加载iris数据集iris = load_iris() X = iris.data y = iris.target#分割数据并X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=4)#建立模型knn = KNeighborsClassifier()#训练模型knn.fit(X_train, y_train)#将准确率打印出print(knn.score(X_test, y_test))# 0.973684210526
Model 交叉验证法(Cross Validation):
from sklearn.cross_validation import cross_val_score # K折交叉验证模块#使用K折交叉验证模块scores = cross_val_score(knn, X, y, cv=5, scoring='accuracy')#将5次的预测准确率打印出print(scores)# [ 0.96666667 1. 0.93333333 0.96666667 1. ]#将5次的预测准确平均率打印出print(scores.mean())# 0.973333333333
以准确率(accuracy)判断
import matplotlib.pyplot as plt #可视化模块#建立测试参数集k_range = range(1, 31)
k_scores = []#藉由迭代的方式来计算不同参数对模型的影响，并返回交叉验证后的平均准确率for k in k_range:
knn = KNeighborsClassifier(n_neighbors=k)
scores = cross_val_score(knn, X, y, cv=10, scoring='accuracy')
k_scores.append(scores.mean())#可视化数据plt.plot(k_range, k_scores)
plt.xlabel('Value of K for KNN')
plt.ylabel('Cross-Validated Accuracy')
plt.show()
img1
以平均方差(Mean squared error):
import matplotlib.pyplot as plt
k_range = range(1, 31)
k_scores = []for k in k_range:
knn = KNeighborsClassifier(n_neighbors=k)
loss = -cross_val_score(knn, X, y, cv=10, scoring='mean_squared_error')
k_scores.append(loss.mean())
plt.plot(k_range, k_scores)
plt.xlabel('Value of K for KNN')
plt.ylabel('Cross-Validated MSE')
plt.show()
img2
交叉验证 2 Cross-validation
sklearn.learning_curve: 检视过拟合
Learning curve 检视过拟合:
from sklearn.learning_curve import learning_curve #学习曲线模块from sklearn.datasets import load_digits #digits数据集from sklearn.svm import SVC #Support Vector Classifierimport matplotlib.pyplot as plt #可视化模块import numpy as np
digits = load_digits()
X = digits.data
y = digits.target
train_sizes, train_loss, test_loss = learning_curve(
SVC(gamma=0.001), X, y, cv=10, scoring='mean_squared_error',
train_sizes=[0.1, 0.25, 0.5, 0.75, 1])#平均每一轮所得到的平均方差(共5轮，分别为样本10%、25%、50%、75%、100%)train_loss_mean = -np.mean(train_loss, axis=1)
test_loss_mean = -np.mean(test_loss, axis=1)
plt.plot(train_sizes, train_loss_mean, 'o-', color="r",
label="Training")
plt.plot(train_sizes, test_loss_mean, 'o-', color="g",
label="Cross-validation")
plt.xlabel("Training examples")
plt.ylabel("Loss")
plt.legend(loc="best")
plt.show()
img1
交叉验证 3 Cross-validation
sklearn.validation_curve:检视过拟合
validation_curve 检视过拟合：
from sklearn.learning_curve import validation_curve #validation_curve模块from sklearn.datasets import load_digits
from sklearn.svm import SVC
import matplotlib.pyplot as plt
import numpy as np#digits数据集digits = load_digits()
X = digits.data
y = digits.target#建立参数测试集param_range = np.logspace(-6, -2.3, 5)#使用validation_curve快速找出参数对模型的影响train_loss, test_loss = validation_curve(
SVC(), X, y, param_name='gamma', param_range=param_range, cv=10, scoring='mean_squared_error')#平均每一轮的平均方差train_loss_mean = -np.mean(train_loss, axis=1)
test_loss_mean = -np.mean(test_loss, axis=1)#可视化图形plt.plot(param_range, train_loss_mean, 'o-', color="r",
label="Training")
plt.plot(param_range, test_loss_mean, 'o-', color="g",
label="Cross-validation")
plt.xlabel("gamma")
plt.ylabel("Loss")
plt.legend(loc="best")
plt.show()
img1
保存模型
使用 pickle 保存
使用 joblib 保存
使用 pickle 保存:
from sklearn import svmfrom sklearn import datasets
clf = svm.SVC()
iris = datasets.load_iris()
X, y = iris.data, iris.target
clf.fit(X,y)import pickle #pickle模块#保存Model(注:save文件夹要预先建立，否则会报错)with open('save/clf.pickle', 'wb') as f:
pickle.dump(clf, f)#读取Modelwith open('save/clf.pickle', 'rb') as f:
clf2 = pickle.load(f) #测试读取后的Model
print(clf2.predict(X[0:1]))# [0]使用 joblib 保存:
from sklearn.externals import joblib #jbolib模块#保存Model(注:save文件夹要预先建立，否则会报错)joblib.dump(clf, 'save/clf.pkl')#读取Modelclf3 = joblib.load('save/clf.pkl')#测试读取后的Modelprint(clf3.predict(X[0:1]))# [0]
作者：CrazyWolf_081c
链接：https://www.jianshu.com/p/74f059916e00
随时随地看视频
