Different chunking methods
A chunk represents a single data point in the monitoring results. Recall that there are three methods for chunking your data: based on time, size, or the number of chunks.
In this exercise, you will chunk and visualize the results of the CBPE algorithm for the US Census dataset using size-based and number-based chunking methods.
The nannyml
library is already imported.
Cet exercice fait partie du cours
Monitoring Machine Learning in Python
Exercice interactif pratique
Essayez cet exercice en complétant cet exemple de code.
reference, analysis, analysis_gt = ____.____()
# Initialize the CBPE algorithm
cbpe = nannyml.CBPE(
y_pred_proba='predicted_probability',
y_pred='prediction',
y_true='employed',
metrics = ['roc_auc', 'accuracy'],
problem_type = 'classification_binary',
____ = ____,
)
cbpe = cbpe.fit(reference)
estimated_results = cbpe.estimate(analysis)
estimated_results.plot().show()