Counterfactual explanations—answering the question “what would need to change for a different outcome?”—are among the most powerful tools in the Explainable AI toolbox. In this talk, we go beyond conventional methods and explore how genetic algorithms can be used to generate meaningful, actionable counterfactuals that provide insight into both data and the behavior of machine learning models.

Drawing from real-world scenarios and code examples from the German Credit Risk dataset, we’ll walk through how to:
* Use genetic algorithms to search for minimal, plausible input changes that flip model predictions.
* Evaluate and constrain counterfactuals for realism and interpretability.
* Detect potential model flaws and dataset biases through “what-if” scenario analysis.

Key Takeaways:
* Learn how to use genetic algorithms to generate counterfactual explanations that enhance model transparency and user trust.
* Discover how “what-if” analysis can reveal model weaknesses, dataset flaws, and actionable paths to desired outcomes.
* Gain practical insights into integrating counterfactuals into your AI workflows, with examples grounded in real-world data and Python code.

Audience:
This session is ideal for data scientists, ML practitioners, and AI educators interested in the intersection of optimization and explainability. Whether you're building transparent models or auditing black-box decisions, you’ll leave with tools and strategies to evolve your explanations — literally.

Level: Intermediate

Keywords: Explainable AI, Counterfactuals, Genetic Algorithms, Model Interpretability, Python, Optimization

Based on my book 'Hands-On Genetic Algorithms with Python', 2nd edition