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.

Using surrogate models for global explanations

Using surrogate models for global explanations - 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

Using surrogate models for global explanations

“

- [Instructor] A surrogate model is simply a substitute. When you feel that your project is best served by a black box model, you can build a second model that explains the first. Here's the key difference. Your target variable changes. Your new target variable is the output of the first model. In his excellent book, "Interpretable Machine Learning: A Guide for Making Black Box Models Explainable," which I highly recommend by the way, Christoff Molnar outlines the generic steps in more detail than I will here. We will do the bare minimum at first. A bit later, we'll learn that KNIME can help us with surrogate models. So the first step is to get the predictions from the black box model, then train an interpretable model on the predictions, meaning that the predictions of first model become the dependent variable of the second model, then measure how well it replicates. When we do the demonstration in KNIME,…

Contents