Generative causal explanations of black-box classifiers
arXiv:2006.13913
Abstract
We develop a method for generating causal post-hoc explanations of black-box classifiers based on a learned low-dimensional representation of the data. The explanation is causal in the sense that changing learned latent factors produces a change in the classifier output statistics. To construct these explanations, we design a learning framework that leverages a generative model and information-theoretic measures of causal influence. Our objective function encourages both the generative model to faithfully represent the data distribution and the latent factors to have a large causal influence on the classifier output. Our method learns both global and local explanations, is compatible with any classifier that admits class probabilities and a gradient, and does not require labeled attributes or knowledge of causal structure. Using carefully controlled test cases, we provide intuition that illuminates the function of our objective. We then demonstrate the practical utility of our method on image recognition tasks.
arXiv commentCamera-ready version to appear at NeurIPS 2020
References in corpus (32)
- Adam: A Method for Stochastic Optimization
- Auto-Encoding Variational Bayes
- Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
- Axiomatic Attribution for Deep Networks
- Learning Important Features Through Propagating Activation Differences
- Understanding Black-box Predictions via Influence Functions
- Explaining Machine Learning Classifiers through Diverse Counterfactual Explanations
- Counterfactual Fairness
- Deep Convolutional Inverse Graphics Network
- Disentangling by Factorising
- Towards Robust Interpretability with Self-Explaining Neural Networks
- A Benchmark for Interpretability Methods in Deep Neural Networks
- Towards a Definition of Disentangled Representations
- What Clinicians Want: Contextualizing Explainable Machine Learning for Clinical End Use
- Deep Learning for Case-Based Reasoning through Prototypes: A Neural Network that Explains Its Predictions
- Multi-Level Variational Autoencoder: Learning Disentangled Representations from Grouped Observations
- Learning Deep Disentangled Embeddings with the F-Statistic Loss
- Explaining Classifiers with Causal Concept Effect (CaCE)
- Interpretability via Model Extraction
- Explaining Image Classifiers by Counterfactual Generation
- CXPlain: Causal Explanations for Model Interpretation under Uncertainty
- Interpreting Black Box Predictions using Fisher Kernels
- Causal Learning and Explanation of Deep Neural Networks via Autoencoded Activations
- Variational Information Maximization for Feature Selection
- PC-Fairness: A Unified Framework for Measuring Causality-based Fairness
- A Game Theoretic Approach to Class-wise Selective Rationalization
- Interpreting Neural Network Judgments via Minimal, Stable, and Symbolic Corrections
- Explaining Deep Learning Models - A Bayesian Non-parametric Approach
- Explaining Deep Learning Models using Causal Inference
- Causal Interpretability for Machine Learning -- Problems, Methods and Evaluation
- Learning Interpretable Models with Causal Guarantees
- Explaining Visual Models by Causal Attribution