Interpreting Black-Box Classifiers Using Instance-Level Visual Explanations

By Paolo Tamagnini, Josua Krause, Aritra Dasgupta, Enrico Bertini. 2017

74 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Longo Luca
2. Giulia Vilone
3. Yongsu Ahn
4. Yu-Ru Lin
5. Enrico Bertini
6. Steffen Holter
7. Oscar Gomez
8. Jun Yuan

In Computer Vision and Pattern Recognition

1. Yongsu Ahn
2. Yu-Ru Lin
3. Enrico Bertini
4. Aritra Dasgupta
5. Steffen Holter
6. Oscar Gomez
7. Jun Yuan
8. Josua Krause

Citations in the corpus

• FairSight: Visual Analytics for Fairness in Decision Making (2019, methodology): “Examples include RuleMatrix [29]’s rule-based visual interface for explaining decision rules based on the secondary decision tree model, the distributionbased visual representation for global-level explanation, and Rivelo’s instance- and feature-level representation [38].”
• Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI) (2018, background): “[158] proposed Rivelo, a pedagogical visual analytics interface that enables expert user of binary classifiers by interactively exploring a set of instance-level explanations.”
• ViCE: visual counterfactual explanations for machine learning models (2020, methodology): “Likewise, Rivelo [20] and the Workflow for Visual Diagnostics proposed by Krause et al.”
• Explaining Explanations in AI (2018, neutral): “Visualisation techniques can visually demonstrate the relative influence of features or particular pixels (e.g. in the case of an image classifier), or provide an interface for users to explore textual or visual explanations (Poulin et al., 2006; Tamagnini et al., 2017).”
• A Workflow for Visual Diagnostics of Binary Classifiers using Instance-Level Explanations (2017, methodology): “Also, contrary to our previous approach of using explanations to detect only the commonly used features by a model [30], here we focus on explanations as a way for experts to diagnose correct or problematic model behavior and address the goals G1, G2, and G3 , that were outlined in Section 3.”
• Explainable Artificial Intelligence: a Systematic Review (2020, background): “Rivelo works exclusively with binary classification problems and binary input features.”