Perfectly parallel fairness certification of neural networks

Recently, there is growing concern that machine-learned software, which currently assists or even automates decision making, reproduces, and in the worst case reinforces, bias present in the training data. The development of tools and techniques for certifying fairness of this software or describing...

Full description

Saved in:
Bibliographic Details
Published in:Proceedings of ACM on programming languages 2020-11, Vol.4 (OOPSLA), p.1-30, Article 185
Main Authors: Urban, Caterina, Christakis, Maria, Wüstholz, Valentin, Zhang, Fuyuan
Format: Article
Language:English
Subjects:
Abstraction
Computer Science
Computers and Society
Computing / technology policy
Computing methodologies
Formal software verification
Logic in Computer Science
Machine Learning
Machine learning approaches
Neural networks
Program analysis
Program reasoning
Programming Languages
Semantics and reasoning
Social and professional topics
Software and its engineering
Software creation and management
Software verification and validation
Theory of computation
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Recently, there is growing concern that machine-learned software, which currently assists or even automates decision making, reproduces, and in the worst case reinforces, bias present in the training data. The development of tools and techniques for certifying fairness of this software or describing its biases is, therefore, critical. In this paper, we propose a perfectly parallel static analysis for certifying fairness of feed-forward neural networks used for classification of tabular data. When certification succeeds, our approach provides definite guarantees, otherwise, it describes and quantifies the biased input space regions. We design the analysis to be sound, in practice also exact, and configurable in terms of scalability and precision, thereby enabling pay-as-you-go certification. We implement our approach in an open-source tool called Libra and demonstrate its effectiveness on neural networks trained on popular datasets.
ISSN:2475-1421
2475-1421
DOI:10.1145/3428253