具有分层交叉验证的多个性能指标

Posted 2023-03-12

【中文标题】具有分层交叉验证的多个性能指标

【英文标题】：Multiple performance metrics with stratified cross validation

【发布时间】：2018-09-19 21:45:18

【问题描述】：

我有一个小的、不平衡的数据集，我想用不同的算法对其进行测试。出于评估目的，我需要多个性能指标（准确度、精确度、召回率、fscore、支持）。

我打算这样做，但我并不满意，因为可能有更简单的解决方案：

skf = StratifiedKFold(n_splits=3, random_state=42, shuffle=True)
accuracy = []
for train_index, test_index in skf.split(X,Y):
    X_train, X_test = X[train_index], X[test_index]
    y_train, y_test = Y[train_index], Y[test_index]
    gradientBoost.fit(X_train, y_train)
    y_pred = gradientBoost.predict(X_test)

    accuracy.append(round(accuracy_score(y_test, y_pred), 2))
    precision, recall, fscore, support = np.round(score(y_test, y_pred), 2)

    print('precision: ' + str(precision))
    print('recall: ' + str(recall))
    print('fscore: ' + str(fscore))
    print('support: ' + str(support))

    print(classification_report(y_test, y_pred))

meanAcc= np.mean(np.asarray(accuracy))
print('meanAcc: ', meanAcc)

理论上，我可以对所有指标进行平均，就像我为准确性所做的那样。有没有更简单和/或更有效的方法？

编辑：

我尝试将准确率和召回加权作为得分手。不幸的是，情节中只显示了准确性。在图例中提到了准确率+召回率。

#Initialize classifier
clf_gini = DecisionTreeClassifier(criterion = "gini", random_state = 42,
                               max_depth=10, min_samples_leaf=8)

scoring = 'Accuracy' : make_scorer(accuracy_score), 'Recall' : 'recall_weighted'

gs = GridSearchCV(DecisionTreeClassifier(criterion= 'entropy', random_state=42, min_samples_leaf = 10), param_grid='max_depth' : range(2, 30, 2),
                  scoring=scoring, cv=3, refit='Accuracy')

gs.fit(X_Distances, Y)
results = gs.cv_results_

plt.figure(figsize=(13, 13))
plt.title("GridSearchCV evaluating using multiple scorers simultaneously",
          fontsize=16)

plt.xlabel("max_depth")
plt.ylabel("Score")
plt.grid()

ax = plt.axes()
ax.set_xlim(0, 32)
ax.set_ylim(0, 1)

# Get the regular numpy array from the MaskedArray
X_axis = np.array(results['param_max_depth'].data, dtype=float)

for scorer, color in zip(sorted(scoring), ['g', 'k']):
    for sample, style in (('train', '--'), ('test', '-')):
        sample_score_mean = results['mean_%s_%s' % (sample, scorer)]
        sample_score_std = results['std_%s_%s' % (sample, scorer)]
        ax.fill_between(X_axis, sample_score_mean - sample_score_std,
                        sample_score_mean + sample_score_std,
                        alpha=0.1 if sample == 'test' else 0, color=color)
        ax.plot(X_axis, sample_score_mean, style, color=color,
                alpha=1 if sample == 'test' else 0.7,
                label="%s (%s)" % (scorer, sample))

        best_index = np.nonzero(results['rank_test_%s' % scorer] == 1)[0][0]
        best_score = results['mean_test_%s' % scorer][best_index]

        # Plot a dotted vertical line at the best score for that scorer marked by x
        ax.plot([X_axis[best_index], ] * 2, [0, best_score],
                linestyle='-.', color=color, marker='x', markeredgewidth=3, ms=8)


    # Annotate the best score for that scorer
    ax.annotate("%0.2f" % best_score,
                (X_axis[best_index], best_score + 0.005))

plt.legend(loc="best")
plt.grid('off')
plt.show()

【问题讨论】：

为什么不附加到列表或字典本身？我不确定你还能做什么。

附加到列表是什么意思？对于我目前附加到列表的准确性，我可以对所有指标都这样做，是的。

是的，将它们附加到列表中，就像您正在做的那样以确保准确性。

【参考方案1】：

我们可以使用GridSearchCV for multi-metric evaluation:

# Author: Raghav RV <[e-mail]>
# License: BSD
import numpy as np
from matplotlib import pyplot as plt

from sklearn.datasets import make_hastie_10_2
from sklearn.model_selection import GridSearchCV
from sklearn.metrics import make_scorer
from sklearn.metrics import accuracy_score
from sklearn.tree import DecisionTreeClassifier

使用多个评估指标运行 GridSearchCV¶

X, y = make_hastie_10_2(n_samples=8000, random_state=42)

# The scorers can be either be one of the predefined metric strings or a scorer
# callable, like the one returned by make_scorer
scoring = 'AUC': 'roc_auc', 'Accuracy': make_scorer(accuracy_score)

# Setting refit='AUC', refits an estimator on the whole dataset with the
# parameter setting that has the best cross-validated AUC score.
# That estimator is made available at ``gs.best_estimator_`` along with
# parameters like ``gs.best_score_``, ``gs.best_parameters_`` and
# ``gs.best_index_``
gs = GridSearchCV(DecisionTreeClassifier(random_state=42),
                  param_grid='min_samples_split': range(2, 403, 10),
                  scoring=scoring, cv=5, refit='AUC')
gs.fit(X, y)
results = gs.cv_results_

绘制结果

plt.figure(figsize=(13, 13))
plt.title("GridSearchCV evaluating using multiple scorers simultaneously",
          fontsize=16)

plt.xlabel("min_samples_split")
plt.ylabel("Score")
plt.grid()

ax = plt.axes()
ax.set_xlim(0, 402)
ax.set_ylim(0.73, 1)

# Get the regular numpy array from the MaskedArray
X_axis = np.array(results['param_min_samples_split'].data, dtype=float)

for scorer, color in zip(sorted(scoring), ['g', 'k']):
    for sample, style in (('train', '--'), ('test', '-')):
        sample_score_mean = results['mean_%s_%s' % (sample, scorer)]
        sample_score_std = results['std_%s_%s' % (sample, scorer)]
        ax.fill_between(X_axis, sample_score_mean - sample_score_std,
                        sample_score_mean + sample_score_std,
                        alpha=0.1 if sample == 'test' else 0, color=color)
        ax.plot(X_axis, sample_score_mean, style, color=color,
                alpha=1 if sample == 'test' else 0.7,
                label="%s (%s)" % (scorer, sample))

        best_index = np.nonzero(results['rank_test_%s' % scorer] == 1)[0][0]
        best_score = results['mean_test_%s' % scorer][best_index]

        # Plot a dotted vertical line at the best score for that scorer marked by x
        ax.plot([X_axis[best_index], ] * 2, [0, best_score],
                linestyle='-.', color=color, marker='x', markeredgewidth=3, ms=8)


    # Annotate the best score for that scorer
    ax.annotate("%0.2f" % best_score,
                (X_axis[best_index], best_score + 0.005))

plt.legend(loc="best")
plt.grid('off')
plt.show()

结果：

【讨论】：

看起来不错。我有一个多类问题，所以我选择了recall_weighted 和accuracy 作为记分器。不幸的是，只绘制了准确性，有什么想法吗？ （我用我的代码更新了我的初始帖子）

【参考方案2】：

sklearn 文档建议使用以下指标之一来评估分类：

accuracy metrics.accuracy_score average_precision metrics.average_precision_score 二元目标的 f1 metrics.f1_score f1_micro metrics.f1_score 微平均 f1_macro metrics.f1_score 宏观平均 f1_weighted metrics.f1_score 加权平均 f1_samples metrics.f1_score by multilabel sample neg_log_loss metrics.log_loss 需要 predict_proba 支持 precision metrics.precision_score 后缀适用于 f1 recall metrics.recall_score 后缀适用于 f1 roc_auc metrics.roc_auc_score

让我们试试accuracy 和f1_weighted：

from sklearn.model_selection import GridSearchCV
from sklearn.datasets import make_classification
from sklearn.metrics import recall_score, make_scorer, accuracy_score
from sklearn.ensemble import RandomForestClassifier


X, y = make_classification(n_classes=10, n_informative=8, random_state=1)

scoring = 
  'Accuracy' : 'accuracy',
  'F1 (macro)' : 'f1_weighted',


scoring = 
  'Accuracy' : 'accuracy',
  'Recall' : 'f1_weighted',


gs = GridSearchCV(RandomForestClassifier(max_depth=5, random_state=42, min_samples_leaf = 10),
                  param_grid='n_estimators' : range(2, 101, 2), return_train_score=True,
                  scoring=scoring, cv=3, refit='Accuracy')

gs.fit(X, y)
results = gs.cv_results_

##################
plt.figure(figsize=(12, 8))
plt.title("GridSearchCV evaluating using multiple scorers simultaneously",
          fontsize=16)

plt.xlabel("n_estimators")
plt.ylabel("Score")
#plt.grid()

ax = plt.gca()
ax.set_xlim(0, 101)
ax.set_ylim(0, 1)

# Get the regular numpy array from the MaskedArray
X_axis = np.array(results['param_n_estimators'].data, dtype=float)

for scorer, color in zip(sorted(scoring), ['g', 'k']):
    for sample, style in (('train', '--'), ('test', '-')):
        print('plotting:  ()'.format(scorer, sample))
        sample_score_mean = results['mean_%s_%s' % (sample, scorer)]
        sample_score_std = results['std_%s_%s' % (sample, scorer)]
        ax.fill_between(X_axis, sample_score_mean - sample_score_std,
                        sample_score_mean + sample_score_std,
                        alpha=0.1 if sample == 'test' else 0, color=color)
        ax.plot(X_axis, sample_score_mean, style, color=color,
                alpha=1 if sample == 'test' else 0.7,
                label="%s (%s)" % (scorer, sample))

        best_index = np.nonzero(results['rank_test_%s' % scorer] == 1)[0][0]
        best_score = results['mean_test_%s' % scorer][best_index]

        # Plot a dotted vertical line at the best score for that scorer marked by x
        ax.plot([X_axis[best_index], ] * 2, [0, best_score],
                linestyle='-.', color=color, marker='x', markeredgewidth=3, ms=8)


    # Annotate the best score for that scorer
    ax.annotate("%0.2f" % best_score,
                (X_axis[best_index], best_score + 0.005))

plt.legend(loc="best")

plt.grid(False)
plt.show()

结果：

【讨论】：

谢谢，几分钟前我发现我只是认为没有绘制附加指标，因为例如，recall_weighted 是与准确度相同的曲线。当我切换到宏时，它是可见的。

@user3667018，是的，我昨天也得出了同样的结论。我应该在答案中提到它......

好的，非常感谢！还有一个问题：将我的整个数据集传递给GridSearchCV(cv=3) 是否合理？我只是在考虑过拟合等问题，但GridSearchCV() 应该处理交叉验证吗？

@user3667018，不行，你的模型在训练过程中应该不能看到测试数据集，否则会出现“数据泄露”...

当我想最终进行分层交叉验证时，我该如何处理？我没有一个单一的训练/测试集。参数调整后，我正在使用找到的最佳参数训练一个新模型并执行交叉验证。