Causal Interpretability for Machine Learning -- Problems, Methods and Evaluation
arXiv:2003.03934
Abstract
Machine learning models have had discernible achievements in a myriad of applications. However, most of these models are black-boxes, and it is obscure how the decisions are made by them. This makes the models unreliable and untrustworthy. To provide insights into the decision making processes of these models, a variety of traditional interpretable models have been proposed. Moreover, to generate more human-friendly explanations, recent work on interpretability tries to answer questions related to causality such as "Why does this model makes such decisions?" or "Was it a specific feature that caused the decision made by the model?". In this work, models that aim to answer causal questions are referred to as causal interpretable models. The existing surveys have covered concepts and methodologies of traditional interpretability. In this work, we present a comprehensive survey on causal interpretable models from the aspects of the problems and methods. In addition, this survey provides in-depth insights into the existing evaluation metrics for measuring interpretability, which can help practitioners understand for what scenarios each evaluation metric is suitable.
References in corpus (30)
- Distilling the Knowledge in a Neural Network
- Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
- Towards A Rigorous Science of Interpretable Machine Learning
- Striving for Simplicity: The All Convolutional Net
- European Union regulations on algorithmic decision-making and a "right to explanation"
- Understanding Black-box Predictions via Influence Functions
- Explaining Machine Learning Classifiers through Diverse Counterfactual Explanations
- Counterfactual Fairness
- Model-Agnostic Interpretability of Machine Learning
- Sanity Checks for Saliency Maps
- Distilling a Neural Network Into a Soft Decision Tree
- Model Agnostic Supervised Local Explanations
- DeepFool: a simple and accurate method to fool deep neural networks
- Explaining Classifiers with Causal Concept Effect (CaCE)
- Interpretable Credit Application Predictions With Counterfactual Explanations
- On the Accuracy of Influence Functions for Measuring Group Effects
- The Limitations of Deep Learning in Adversarial Settings
- Generating Counterfactual and Contrastive Explanations using SHAP
- Generating Counterfactual Explanations with Natural Language
- Explaining Latent Factor Models for Recommendation with Influence Functions
- Causal Learning and Explanation of Deep Neural Networks via Autoencoded Activations
- Evaluating Explanation Without Ground Truth in Interpretable Machine Learning
- Neural Network Attributions: A Causal Perspective
- Generative Counterfactual Introspection for Explainable Deep Learning
- Counterfactuals uncover the modular structure of deep generative models
- Explaining Deep Learning Models using Causal Inference
- Learning Interpretable Models with Causal Guarantees
- Explaining Visual Models by Causal Attribution
- Fairness Through Causal Awareness: Learning Latent-Variable Models for Biased Data
- Multimodal Explanations by Predicting Counterfactuality in Videos