Interpretability via Model Extraction
arXiv:1706.09773
Abstract
The ability to interpret machine learning models has become increasingly important now that machine learning is used to inform consequential decisions. We propose an approach called model extraction for interpreting complex, blackbox models. Our approach approximates the complex model using a much more interpretable model; as long as the approximation quality is good, then statistical properties of the complex model are reflected in the interpretable model. We show how model extraction can be used to understand and debug random forests and neural nets trained on several datasets from the UCI Machine Learning Repository, as well as control policies learned for several classical reinforcement learning problems.
arXiv commentPresented as a poster at the 2017 Workshop on Fairness, Accountability, and Transparency in Machine Learning (FAT/ML 2017)
References in corpus (3)
Cited by in corpus (31)
- Explaining Explanations in AI
- Verifiable Reinforcement Learning via Policy Extraction
- Explainable Artificial Intelligence: a Systematic Review
- Tree-based local explanations of machine learning model predictions, AraucanaXAI
- Fairness via Explanation Quality: Evaluating Disparities in the Quality of Post hoc Explanations
- Testing and verification of neural-network-based safety-critical control software: A systematic literature review
- On Interpretability of Artificial Neural Networks: A Survey
- A survey of algorithmic recourse: definitions, formulations, solutions, and prospects
- A framework for the extraction of Deep Neural Networks by leveraging public data
- Techniques for Interpretable Machine Learning
- "How do I fool you?": Manipulating User Trust via Misleading Black Box Explanations
- The RFML Ecosystem: A Look at the Unique Challenges of Applying Deep Learning to Radio Frequency Applications
- Born-Again Tree Ensembles
- Beyond Individualized Recourse: Interpretable and Interactive Summaries of Actionable Recourses
- Reliable Post hoc Explanations: Modeling Uncertainty in Explainability
- Towards Interpretable-AI Policies Induction using Evolutionary Nonlinear Decision Trees for Discrete Action Systems
- Counterfactuals and Causability in Explainable Artificial Intelligence: Theory, Algorithms, and Applications
- Interpretable Deep Learning under Fire
- Interpreting Deep Classifier by Visual Distillation of Dark Knowledge
- Generative causal explanations of black-box classifiers
- Why model why? Assessing the strengths and limitations of LIME
- i-Align: an interpretable knowledge graph alignment model
- Provably efficient, succinct, and precise explanations
- Algorithmic Recourse in the Wild: Understanding the Impact of Data and Model Shifts
- Robust and Stable Black Box Explanations
- Towards the Unification and Robustness of Perturbation and Gradient Based Explanations
- M2Lens: Visualizing and Explaining Multimodal Models for Sentiment Analysis
- BODAME: Bilevel Optimization for Defense Against Model Extraction
- Designing Interpretable Approximations to Deep Reinforcement Learning
- Human-grounded Evaluations of Explanation Methods for Text Classification
- Interpretability of Blackbox Machine Learning Models through Dataview Extraction and Shadow Model creation