InfoGram and Admissible Machine Learning
–arXiv.org Artificial Intelligence
We have entered a new era of machine learning (ML), where the most accurate algorithm with superior predictive power may not even be deployable, unless it is admissible under the regulatory constraints. This has led to great interest in developing fair, transparent and trustworthy ML methods. The purpose of this article is to introduce a new information-theoretic learning framework (admissible machine learning) and algorithmic risk-management tools (InfoGram, L-features, ALFA-testing) that can guide an analyst to redesign off-the-shelf ML methods to be regulatory compliant, while maintaining good prediction accuracy. We have illustrated our approach using several real-data examples from financial sectors, biomedical research, marketing campaigns, and the criminal justice system.
arXiv.org Artificial Intelligence
Aug-19-2021
- Country:
- Europe (0.14)
- Asia > Taiwan (0.04)
- North America > United States
- Wisconsin (0.04)
- New York (0.04)
- Florida > Broward County (0.04)
- California > Santa Clara County
- Palo Alto (0.04)
- Genre:
- Research Report > Experimental Study (0.47)
- Industry:
- Law > Criminal Law (1.00)
- Banking & Finance > Credit (1.00)
- Law Enforcement & Public Safety > Crime Prevention & Enforcement (0.92)
- Health & Medicine
- Therapeutic Area > Oncology (1.00)
- Pharmaceuticals & Biotechnology (1.00)
- Nuclear Medicine (0.67)
- Diagnostic Medicine > Imaging (0.67)
- Government > Regional Government
- Technology: