Explanations of model predictions with live and breakDown packages

Abstract

Complex models are commonly used in predictive modeling. In this paper we present R packages that can be used to explain predictions from complex black box models and attribute parts of these predictions to input features. We introduce two new approaches and corresponding packages for such attribution, namely live and breakDown. We also compare their results with existing implementations of state of the art solutions, namely lime that implements Locally Interpretable Model-agnostic Explanations and ShapleyR that implements Shapley values.

References in corpus (6)

• A Unified Approach to Interpreting Model Predictions
• Visualizing the Effects of Predictor Variables in Black Box Supervised Learning Models
• Nothing Else Matters: Model-Agnostic Explanations By Identifying Prediction Invariance
• MAGIX: Model Agnostic Globally Interpretable Explanations
• archivist: An R Package for Managing, Recording and Restoring Data Analysis Results
• DALEX: explainers for complex predictive models

Cited by in corpus (3)

• Phase Classification of Multi-Principal Element Alloys via Interpretable Machine Learning
• The Grammar of Interactive Explanatory Model Analysis
• Towards Explainable Test Case Prioritisation with Learning-to-Rank Models