机器学习第3章分类

当前位置:

首页 > 编程开发 > 数据分析 >

机器学习第3章分类

机器学习实战：基于Scikit-Learn和TensorFlow的笔记

参考：作者的Jupyter Notebook
Chapter 2 – End-to-end Machine Learning project

获取MNIST数据集的代码：

def sort_by_target(mnist):
    reorder_train = np.array(sorted([(target, i) for i, target in enumerate(mnist.target[:60000])]))[:, 1]
    reorder_test = np.array(sorted([(target, i) for i, target in enumerate(mnist.target[60000:])]))[:, 1]
    mnist.data[:60000] = mnist.data[reorder_train]
    mnist.target[:60000] = mnist.target[reorder_train]
    mnist.data[60000:] = mnist.data[reorder_test + 60000]
    mnist.target[60000:] = mnist.target[reorder_test + 60000]
from sklearn.datasets import fetch_openml
mnist = fetch_openml('mnist_784', version=1, cache=True)
mnist.target = mnist.target.astype(np.int8) # fetch_openml() returns targets as strings
sort_by_target(mnist) # fetch_openml() returns an unsorted dataset

查看这些数组

#print(mnist["data"], mnist["target"])
#print(mnist.data.shape)
X, y = mnist["data"], mnist["target"]
#print(X.shape)
#print(y.shape)

some_digit = X[36000]
some_digit_image = some_digit.reshape(28, 28)
plt.imshow(some_digit_image, cmap = mpl.cm.binary,
        interpolation="nearest")
plt.axis("off")
#plt.show()
#print(y[36000])

MNIST数据集中的部分数字图像

def plot_digits(instances, images_per_row=10, **options):
    size = 28
    images_per_row = min(len(instances), images_per_row)
    images = [instance.reshape(size,size) for instance in instances]
    n_rows = (len(instances) - 1) // images_per_row + 1
    row_images = []
    n_empty = n_rows * images_per_row - len(instances)
    images.append(np.zeros((size, size * n_empty)))
    for row in range(n_rows):
        rimages = images[row * images_per_row : (row + 1) * images_per_row]
        row_images.append(np.concatenate(rimages, axis=1))
    image = np.concatenate(row_images, axis=0)
    plt.imshow(image, cmap = mpl.cm.binary, **options)
    plt.axis("off")

plt.figure(figsize=(9,9))
example_images = np.r_[X[:12000:600], X[13000:30600:600], X[30600:60000:590]]
plot_digits(example_images, images_per_row=10)
#save_fig("more_digits_plot")
#plt.show()

给数据集洗牌

X_train, X_test, y_train, y_test = X[:60000], X[60000:], y[:60000], y[60000:]
shuffle_index = np.random.permutation(60000)
X_train, y_train = X_train[shuffle_index], y_train[shuffle_index]

训练一个二元分类器，为此分类任务创建目标向量：

y_train_5 = (y_train == 5)  # True for all 5s, False for all other digits.
y_test_5 = (y_test == 5)

创建一个SGDClassifier(随机梯度下降分类器)并在整个训练集上进行训练：

from sklearn.linear_model import SGDClassifier
sgd_clf = SGDClassifier(max_iter=5, tol=-np.infty, random_state=42)   #random_state=42
sgd_clf.fit(X_train, y_train_5)
#print(sgd_clf.fit(X_train, y_train_5))
#现在可以用它来检测数字5的图像了：
sgd_clf.predict([some_digit])
#print(sgd_clf.predict([some_digit]))

交叉验证

from sklearn.model_selection import cross_val_score
cross_val_score(sgd_clf, X_train, y_train_5, cv=3, scoring="accuracy")
print(cross_val_score(sgd_clf, X_train, y_train_5, cv=3, scoring="accuracy"))

#下面这段代码与前面的cross_val_score（）大致相同，并打印出相同的结果：
from sklearn.model_selection import StratifiedKFold
from sklearn.base import clone
skfolds = StratifiedKFold(n_splits=3, random_state=42)   #random_state=42
for train_index, test_index in skfolds.split(X_train, y_train_5):
    clone_clf = clone(sgd_clf)
    X_train_folds = X_train[train_index]
    y_train_folds = (y_train_5[train_index])
    X_test_fold = X_train[test_index]
    y_test_fold = (y_train_5[test_index])

    clone_clf.fit(X_train_folds, y_train_folds)
    y_pred = clone_clf.predict(X_test_fold)
    n_correct = sum(y_pred == y_test_fold)
    print(n_correct / len(y_pred))

一个蠢笨的分类器(不是我说的)，它将每张图都分类成“非5”：

from sklearn.base import BaseEstimator
class Never5Classifier(BaseEstimator):
    def fit(self, X, y=None):
        pass
    def predict(self, X):
        return np.zeros((len(X), 1), dtype=bool)
#准确度
never_5_clf = Never5Classifier()
print(cross_val_score(never_5_clf, X_train, y_train_5, cv=3, scoring="accuracy"))

混淆矩阵:评估分类器性能的更好方法是混淆矩阵。

from sklearn.model_selection import cross_val_predict
y_train_pred = cross_val_predict(sgd_clf, X_train, y_train_5, cv=3)
#cross_val_predict（）函数同样执行K-fold交叉验证，但返回的不是评估分数，而是每个折叠的预测。这意味着对于每个实例都可以得到一个干净的预测
from sklearn.metrics import confusion_matrix
#confusion_matrix(y_train_5, y_train_pred)
#print(confusion_matrix(y_train_5, y_train_pred))
y_train_perfect_predictions = y_train_5
#print(confusion_matrix(y_train_5, y_train_perfect_predictions))

精度和召回率

#精度=TP/(TP+FP)：TP是真正类的数量，FP是假正类的数量。
#召回率=TP/(TP+FN)：FN是假负类的数量。
from sklearn.metrics import precision_score, recall_score
print(precision_score(y_train_5, y_train_pred))  #精度4344 / (4344 + 1307)
print(recall_score(y_train_5, y_train_pred))  #召回率4344 / (4344 + 1077)

#F1分数：F1=2/（1/精度+1/召回率）=TP/(TP+(FN+FP)/2)
from sklearn.metrics import f1_score
print(f1_score(y_train_5, y_train_pred))

精度/召回率权衡：阈值

y_scores = sgd_clf.decision_function([some_digit])
print(y_scores)
threshold = 0
y_some_digit_pred = (y_scores > threshold)
print(y_some_digit_pred)
#提高阈值
threshold = 200000
y_some_digit_pred_a = (y_scores > threshold)
print(y_some_digit_pred_a)

决定使用什么阈值

#获取训练集中所有实例的分数
y_scores = cross_val_predict(sgd_clf, X_train, y_train_5, cv=3, method="decision_function")
#计算所有可能的阈值的精度和召回率
from sklearn.metrics import precision_recall_curve
precisions, recalls, thresholds = precision_recall_curve(y_train_5, y_scores)
#使用Matplotlib绘制精度和召回率相对于阈值的函数图
def plot_precision_recall_vs_threshold(precisions, recalls, thresholds):
    plt.plot(thresholds, precisions[:-1], "b--", label="Precision")
    plt.plot(thresholds, recalls[:-1], "g-", label="Recall")
    plt.xlabel("Threshold")
    plt.legend(loc="upper left")
    plt.ylim([0, 1])
plt.figure(figsize=(8, 4))
plot_precision_recall_vs_threshold(precisions, recalls, thresholds)
plt.xlim([-700000, 700000])
plt.show()
#print((y_train_pred == (y_scores > 0)).all())
y_train_pred_90 = (y_scores > 70000)
from sklearn.metrics import precision_score, recall_score
print(precision_score(y_train_5, y_train_pred_90)) #精度
print(recall_score(y_train_5, y_train_pred_90)) #召回率

精度和召回率的函数图PR

def plot_precision_vs_recall(precisions, recalls):
    plt.plot(recalls, precisions, "b-", linewidth=2)
    plt.xlabel("Recall", fontsize=16)
    plt.ylabel("Precision", fontsize=16)
    plt.axis([0, 1, 0, 1])

plt.figure(figsize=(8, 6))
plot_precision_vs_recall(precisions, recalls)
plt.show()

ROC曲线（受试者工作特征曲线）：真正类率和假正类率

from sklearn.metrics import roc_curve
fpr, tpr, thresholds = roc_curve(y_train_5, y_scores)
def plot_roc_curve(fpr, tpr, label=None):
    plt.plot(fpr, tpr, linewidth=2, label=label)
    plt.plot([0, 1], [0, 1], 'k--')
    plt.axis([0, 1, 0, 1])
    plt.xlabel('False Positive Rate', fontsize=16)
    plt.ylabel('True Positive Rate', fontsize=16)
'''
plt.figure(figsize=(8, 6))
plot_roc_curve(fpr, tpr)
plt.show()
from sklearn.metrics import roc_auc_score
print(roc_auc_score(y_train_5, y_scores))

训练一个RandomForestClassifier分类器，并比较它和SGDClassifier分类器的ROC曲线和ROC AUC分数。

from sklearn.ensemble import RandomForestClassifier
forest_clf = RandomForestClassifier(n_estimators=10, random_state=42)
y_probas_forest = cross_val_predict(forest_clf, X_train, y_train_5, cv=3, method="predict_proba")
y_scores_forest = y_probas_forest[:, 1] # score = proba of positive class
fpr_forest, tpr_forest, thresholds_forest = roc_curve(y_train_5,y_scores_forest)
plt.figure(figsize=(8, 6))
plt.plot(fpr, tpr, "b:", linewidth=2, label="SGD")
plot_roc_curve(fpr_forest, tpr_forest, "Random Forest")
plt.legend(loc="lower right", fontsize=16)
plt.show()
from sklearn.metrics import roc_auc_score
print(roc_auc_score(y_train_5, y_scores_forest))

多类别分类器，用SGDClassifier试试

#用SGDClassifier试试：
sgd_clf.fit(X_train, y_train)
sgd_clf.predict([some_digit])
#print(sgd_clf.predict([some_digit]))
some_digit_scores = sgd_clf.decision_function([some_digit])
#print(some_digit_scores)
#print(np.argmax(some_digit_scores))
#print(sgd_clf.classes_)
#print(sgd_clf.classes_[5])

#下面这段代码使用OvO策略，基于SGDClassifier创建了一个多类别分类器：
from sklearn.multiclass import OneVsOneClassifier
ovo_clf = OneVsOneClassifier(SGDClassifier(max_iter=5, tol=-np.infty, random_state=42))
ovo_clf.fit(X_train, y_train)
ovo_clf.predict([some_digit])
len(ovo_clf.estimators_)
#print(ovo_clf.predict([some_digit]))
#print(len(ovo_clf.estimators_))

训练RandomForestClassifier

from sklearn.model_selection import cross_val_score
forest_clf.fit(X_train, y_train)
#print(forest_clf.predict([some_digit]))
#print(forest_clf.predict_proba([some_digit]))  #概率列表
#print(cross_val_score(sgd_clf, X_train, y_train, cv=3, scoring="accuracy"))  #准确率
#将输入进行简单缩放
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
X_train_scaled = scaler.fit_transform(X_train.astype(np.float64))
#print(cross_val_score(sgd_clf, X_train_scaled, y_train, cv=3, scoring="accuracy"))

使用Matplotlib的matshow（）函数来查看混淆矩阵

y_train_pred = cross_val_predict(sgd_clf, X_train_scaled, y_train, cv=3)
conf_mx = confusion_matrix(y_train, y_train_pred)
#print(conf_mx)
#使用Matplotlib的matshow（）函数来查看混淆矩阵的图像表示
#plt.matshow(conf_mx, cmap=plt.cm.gray)
#save_fig("confusion_matrix_plot", tight_layout=False)

你需要将混淆矩阵中的每个值除以相应类别中的图片数量，这样你比较的就是错误率而不是错误的绝对值

row_sums = conf_mx.sum(axis=1, keepdims=True)
norm_conf_mx = conf_mx / row_sums
#用0填充对角线，只保留错误，重新绘制结果：
np.fill_diagonal(norm_conf_mx, 0)
plt.matshow(norm_conf_mx, cmap=plt.cm.gray)
#save_fig("confusion_matrix_errors_plot", tight_layout=False)

看看数字3和数字5的例子：

cl_a, cl_b = 3, 5
X_aa = X_train[(y_train == cl_a) & (y_train_pred == cl_a)]
X_ab = X_train[(y_train == cl_a) & (y_train_pred == cl_b)]
X_ba = X_train[(y_train == cl_b) & (y_train_pred == cl_a)]
X_bb = X_train[(y_train == cl_b) & (y_train_pred == cl_b)]

plt.figure(figsize=(8,8))
plt.subplot(221); plot_digits(X_aa[:25], images_per_row=5)
plt.subplot(222); plot_digits(X_ab[:25], images_per_row=5)
plt.subplot(223); plot_digits(X_ba[:25], images_per_row=5)
plt.subplot(224); plot_digits(X_bb[:25], images_per_row=5)
#save_fig("error_analysis_digits_plot")

多标签分类

#这段代码会创建一个y_multilabel数组，其中包含两个数字图片的目标标签：第一个表示数字是否是大数（7、8、9），第二个表示是否为奇数。
from sklearn.neighbors import KNeighborsClassifier
y_train_large = (y_train >= 7)
y_train_odd = (y_train % 2 == 1)
y_multilabel = np.c_[y_train_large, y_train_odd]

knn_clf = KNeighborsClassifier()
knn_clf.fit(X_train, y_multilabel)
#print(knn_clf.fit(X_train, y_multilabel))
#下一行创建一个KNeighborsClassifier实例（它支持多标签分类，不是所有的分类器都支持），然后使用多个目标数组对它进行
#训练。现在用它做一个预测，注意它输出的两个标签：
knn_clf.predict([some_digit])    #数字5确实不大（False），为奇数（True）。
#print(knn_clf.predict([some_digit]))

下面这段代码计算所有标签的平均F1分数：

from sklearn.metrics import f1_score
y_train_knn_pred = cross_val_predict(knn_clf, X_train, y_multilabel, cv=3, n_jobs=-1)
f1_score(y_multilabel, y_train_knn_pred, average="macro")
#print(f1_score(y_multilabel, y_train_knn_pred, average="macro"))

多输出分类(多输出-多类别分类)

#还先从创建训练集和测试集开始，使用NumPy的randint（）函数
#为MNIST图片的像素强度增加噪声。目标是将图片还原为原始图片：
noise = np.random.randint(0, 100, (len(X_train), 784))
X_train_mod = X_train + noise
noise = np.random.randint(0, 100, (len(X_test), 784))
X_test_mod = X_test + noise
y_train_mod = X_train
y_test_mod = X_test

some_index = 5500
#plt.subplot(121); plot_digit(X_test_mod[some_index])
#plt.subplot(122); plot_digit(y_test_mod[some_index])
#save_fig("noisy_digit_example_plot")

清洗这张图片：

knn_clf.fit(X_train_mod, y_train_mod)
clean_digit = knn_clf.predict([X_test_mod[some_index]])
plot_digit(clean_digit)
save_fig("cleaned_digit_example_plot")
plt.show()

栏目列表