Practical Explainable AI Using Python: Artificial Intelligence Model Explanations Using Python-based Libraries, Extensions, and Frameworks

© The Author(s), under exclusive license to APress Media, LLC, part of Springer Nature 2022

P. MishraPractical Explainable AI Using Pythonhttps://doi.org/10.1007/978-1-4842-7158-2_12

12. Model-Agnostic Explanations by Identifying Prediction Invariance

Pradeepta Mishra¹  

(1)

Sobha Silicon Oasis, Bangalore, Karnataka, India

 

The invariant terminology taken from the field of mathematics explains that predictions generated by a machine learning model remain unchanged if we intervene with the explanatory variables, given the fact that the model is generated through a formal causal relationship. There is a difference between the causal machine learning model and the non-causal machine learning model. The causal model shows a real cause-and-effect relationship ...