SHAP vs. model-specific approaches

You will compare the explanatory power of SHAP values from a Kernel Explainer with the logistic regression coefficients, both trained on the income dataset. A helper function plot_importances() is called at the end of the script to plot importances on the same plot.

X containing the features and y containing the labels, and the logistic regression model have been pre-loaded for you. matplotlib.pyplot has been imported as plt.

This exercise is part of the course

Explainable AI in Python

View Course

Exercise instructions

Compute the coefficients of the logistic regression model.
Create Kernel Explainer to compute shap_values using the logistic regression model and a k-means summary of 10 samples from X.
Compute the mean absolute SHAP values to estimate the impact of each feature.

Hands-on interactive exercise

Have a go at this exercise by completing this sample code.

import shap

# Extract model coefficients
coefficients = ____

# Compute SHAP values
explainer = ____
shap_values = explainer.shap_values(X)

# Calculate mean absolute SHAP values
mean_abs_shap = ____

plot_importances(coefficients, mean_abs_shap)

Edit and Run Code