Synergize AI and Domain Expertise - Explainability Check with Python

Explore the importance of Model Explainability in AI through a 27-minute conference talk from EuroPython 2022. Delve into the Why, How, and What of building consistent, robust, and trustworthy models. Examine the limitations of complex models in delivering meaningful insights and understand cause-effect relationships within data. Learn about explainers and their role in empowering decision-makers beyond mere predictions. Discover the game-theory based algorithm SHAP with a practical Python implementation. Gain insights from two industry applications that highlight the crucial intersections between AI and domain expertise. Cover topics including interpretation vs explainability, correlation vs causation, interconnected effects, local and global explainers, and interaction and dependence in AI models.

Intro
My Experience
Explainability AI
Interpretation vs Explainability
Correlation Causation
Interconnected Effects
Case Study
Model Explainers
Local Explainers
Global Explainers
Interaction and Dependence
Algorithm
Code