Operationalizing the Search for Less Discriminatory Alternatives in Fair Lending
The Less Discriminatory Alternative is a key provision of the disparate impact doctrine in the
United States. In fair lending, this provision mandates that lenders must adopt models that …
United States. In fair lending, this provision mandates that lenders must adopt models that …
Algorithmic Arbitrariness in Content Moderation
Machine learning (ML) is widely used to moderate online content. Despite its scalability
relative to human moderation, the use of ML introduces unique challenges to content …
relative to human moderation, the use of ML introduces unique challenges to content …
Random parameters in learning: advantages and guarantees
E Coupkova - 2024 - hammer.purdue.edu
The generalization error of a classifier is related to the complexity of the set of functions
among which the classifier is chosen. We study a family of low-complexity classifiers …
among which the classifier is chosen. We study a family of low-complexity classifiers …
Natural Learning
H Fanaee-T - arXiv preprint arXiv:2404.05903, 2024 - arxiv.org
We introduce Natural Learning (NL), a novel algorithm that elevates the explainability and
interpretability of machine learning to an extreme level. NL simplifies decisions into intuitive …
interpretability of machine learning to an extreme level. NL simplifies decisions into intuitive …
On the Rashomon ratio of infinite hypothesis sets
E Coupkova, M Boutin - arXiv preprint arXiv:2404.17746, 2024 - arxiv.org
Given a classification problem and a family of classifiers, the Rashomon ratio measures the
proportion of classifiers that yield less than a given loss. Previous work has explored the …
proportion of classifiers that yield less than a given loss. Previous work has explored the …
[PDF][PDF] Interpretability and Multiplicity: a Path to Trustworthy Machine Learning
C Zhong - 2024 - dukespace.lib.duke.edu
Abstract Machine learning has been increasingly deployed for myriad high-stakes decisions
that deeply impact people's lives. This is concerning, because not every model can be …
that deeply impact people's lives. This is concerning, because not every model can be …
The Roads Not Taken: Model Multiplicity in Machine Learning
J Watson-Daniels - 2024 - dash.harvard.edu
In machine learning, model multiplicity is the existence of multiple models that perform
equally well for a given prediction task (also known as the" Rashomon effect"). The set of …
equally well for a given prediction task (also known as the" Rashomon effect"). The set of …
[PDF][PDF] In Pursuit of Simplicity: The Role of the Rashomon Effect for Informed Decision Making
L Semenova - 2024 - dukespace.lib.duke.edu
For high-stakes decision domains, such as healthcare, lending, and criminal justice, the
predictions of deployed models can have a huge impact on human lives. The understanding …
predictions of deployed models can have a huge impact on human lives. The understanding …
Transition Noise Facilitates Interpretability
Recent research in supervised learning has demonstrated that noise in data generation
processes leads to the existence of accurate and simpler/interpretable machine learning …
processes leads to the existence of accurate and simpler/interpretable machine learning …
[PDF][PDF] Amazing Things Come From Having Many Good Models
Abstract The Rashomon Effect, coined by Leo Breiman, describes the phenomenon that
there exist many equally good predictive models for the same dataset. This phenomenon …
there exist many equally good predictive models for the same dataset. This phenomenon …