Kernel explainer for MLPClassifier

Neural networks can be very accurate, but understanding their decisions can be challenging due to complexity. Now, you'll leverage the SHAP Kernel Explainer to interpret an MLPClassifier trained on the adult income dataset. You will explore which of the three features—age, education, or hours worked per week—is most important for predicting income according to this model.

X containing the predictors and y containing the admission decisions, along with the pre-trained MLPClassifier model, have been pre-loaded for you.

This exercise is part of the course

Explainable AI in Python

View Course

Exercise instructions

Instantiate a SHAP Kernel Explainer using the MLPClassifier model and a k-means summary of 10 samples from X.
Generate shap_values for X.
Compute the mean absolute SHAP values to identify key factors affecting admissions.

Hands-on interactive exercise

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

import shap

# Create a SHAP Kernel Explainer
explainer = ____

# Calculate SHAP values
shap_values = ____

# Calculate mean absolute SHAP values
mean_abs_shap = ____

plt.bar(X.columns, mean_abs_shap)
plt.title('Mean Absolute SHAP Values for MLPClassifier')
plt.xticks(rotation=45)
plt.show()

Edit and Run Code