There is a newer version of the record available.

Published April 10, 2024 | Version Demonstrator
Python Library Restricted

Shap

Owner

Description

SHAP is a game theoretic approach to explain the output of any machine learning model. It connects optimal credit allocation with local explanations using the classic Shapley values from game theory and their related extensions

Files

Files

Restricted

The record is publicly accessible, but files are restricted to users with access.

Additional details

Engineering roles
Human Cognition Engineer
Embedded Software Engineer
ML-Algorithm Engineer
Data Engineer
Functional Set
Explainability
Download Link
pip install Shap
Technological maturity
Image Link
data:image/png;base64,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