From the course: Machine Learning and AI Foundations: Producing Explainable AI (XAI) and Interpretable Machine Learning Solutions

Unlock the full course today

Join today to access over 24,900 courses taught by industry experts.

Introducing SHAP

Introducing SHAP - KNIME Tutorial

From the course: Machine Learning and AI Foundations: Producing Explainable AI (XAI) and Interpretable Machine Learning Solutions

Start my 1-month free trial Buy for my team

Introducing SHAP

“

- [Instructor] While Shapley values have been borrowed from game theory, there is a much younger concept based on Shapley values but that was specifically developed for XAI SHAP. First, the acronym, it stands for shapely additive explanations and was proposed by Scott Lundberg and his co-authors. We don't need to get deep into the math. For our purposes, it will be sufficient to review just a couple of high level points. First, an explicit goal of SHAP is to try to perform these calculations faster. The Shapley value algorithm is not very scalable. Next SHAP attempts to explain the delta or difference between a case and a base rate. Remember that we are using SHAP for local explanations. So we're not trying to explain the whole model. So what we care about is explaining how each variable moves a particular case higher or lower than the group. We're trying to answer, for instance, why Anne's credit risk might be lower…

Contents