解释黑箱分类器

在本笔记本中，我们将使用interpret包通过SHAP、Lime、MorrisSensitivity和PartialDependence来解释黑盒分类器。

这个笔记本可以在我们的examples folder在GitHub上找到。

# install interpret if not already installed
try:
    import interpret
except ModuleNotFoundError:
    !pip install --quiet interpret pandas scikit-learn lime

import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split
from interpret import show

from interpret import set_visualize_provider
from interpret.provider import InlineProvider
set_visualize_provider(InlineProvider())

df = pd.read_csv(
    "https://archive.ics.uci.edu/ml/machine-learning-databases/adult/adult.data",
    header=None)
df.columns = [
    "Age", "WorkClass", "fnlwgt", "Education", "EducationNum",
    "MaritalStatus", "Occupation", "Relationship", "Race", "Gender",
    "CapitalGain", "CapitalLoss", "HoursPerWeek", "NativeCountry", "Income"
]
X = df.iloc[:, :-1]
y = (df.iloc[:, -1] == " >50K").astype(int)

# We have to transform categorical variables to use sklearn models
X = pd.get_dummies(X, prefix_sep='.').astype(float)

seed = 42
np.random.seed(seed)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.20, random_state=seed)

训练一个黑箱分类系统

from sklearn.ensemble import RandomForestClassifier
from sklearn.decomposition import PCA
from sklearn.pipeline import Pipeline

#Blackbox system can include preprocessing, not just a classifier!
pca = PCA()
rf = RandomForestClassifier(random_state=seed)

blackbox_model = Pipeline([('pca', pca), ('rf', rf)])
blackbox_model.fit(X_train, y_train)

Pipeline(steps=[('pca', PCA()),
                ('rf', RandomForestClassifier(random_state=42))])

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

显示黑箱模型性能

from interpret.perf import ROC

blackbox_perf = ROC(blackbox_model).explain_perf(X_test, y_test, name='Blackbox')
show(blackbox_perf)

局部解释：单个预测是如何做出的

from interpret.blackbox import LimeTabular

#Blackbox explainers need a predict function, and optionally a dataset
lime = LimeTabular(blackbox_model, X_train, random_state=1)

#Pick the instances to explain, optionally pass in labels if you have them
lime_local = lime.explain_local(X_test[:5], y_test[:5], name='LIME')

show(lime_local, 0)

from interpret.blackbox import ShapKernel

background_val = pd.DataFrame(np.median(X_train, axis=0).reshape(1, -1), columns=X.columns)
shap = ShapKernel(blackbox_model, background_val)
shap_local = shap.explain_local(X_test[:5], y_test[:5], name='SHAP')
show(shap_local, 0)

全局解释：模型整体行为

from interpret.blackbox import MorrisSensitivity

sensitivity = MorrisSensitivity(blackbox_model, X_train)
sensitivity_global = sensitivity.explain_global(name="Global Sensitivity")

show(sensitivity_global)

from interpret.blackbox import PartialDependence

pdp = PartialDependence(blackbox_model, X_train)
pdp_global = pdp.explain_global(name='Partial Dependence')

show(pdp_global, 0)