Explainability requirements in MLOps - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the library used for model explainability.

MLOps

import [1]

Drag options to blanks, or click blank then click option'

Atensorflow

Bshap

Cnumpy

Dpandas

Attempts:

3 left

💡 Hint

Common Mistakes

Importing unrelated libraries like tensorflow or pandas for explainability.

Confusing numpy with explainability tools.

The shap library is commonly used for explainability in machine learning models.

2fill in blank

medium

Complete the code to create an explainer object for a model using SHAP.

MLOps

explainer = shap.[1](model)

Drag options to blanks, or click blank then click option'

ATreeExplainer

Bsummary_plot

Cforce_plot

Ddependence_plot

Attempts:

3 left

💡 Hint

Common Mistakes

Using plot functions instead of explainer constructors.

Confusing summary_plot with explainer creation.

TreeExplainer is used to explain tree-based models in SHAP.

3fill in blank

hard

Fix the error in the code to generate SHAP values for the dataset.

MLOps

shap_values = explainer.[1](X_test)

Drag options to blanks, or click blank then click option'

Aplot

Bshap_values()

Cvalues

Dshap_values

Attempts:

3 left

💡 Hint

Common Mistakes

Trying to call a non-existent method 'values'.

Using 'plot' which is for visualization, not value extraction.

The method to get SHAP values is called shap_values.

4fill in blank

hard

Fill both blanks to create a SHAP summary plot for the test data.

MLOps

shap.[1](shap_values, [2])

Drag options to blanks, or click blank then click option'

Asummary_plot

BX_test

Cforce_plot

Dmodel

Attempts:

3 left

💡 Hint

Common Mistakes

Using force_plot instead of summary_plot for overview.

Passing the model instead of the dataset.

The summary_plot function shows the SHAP values visually, and it needs the SHAP values and the dataset.

5fill in blank

hard

Fill all three blanks to filter features with SHAP values greater than 0.1 and create a dictionary of feature importance.

MLOps

important_features = {feature: value for feature, value in zip([1], shap_values[0]) if value [2] [3]

Drag options to blanks, or click blank then click option'

AX_test.columns

B>

C0.1

DX_test.index

Attempts:

3 left

💡 Hint

Common Mistakes

Using index instead of columns for feature names.

Using wrong comparison operators like '<' or '=='

Using a threshold other than 0.1.

This code creates a dictionary of features with SHAP values greater than 0.1 using the dataset's column names.