Search Results for author:
Found 9 papers, 1 papers with code
Task-specific experimental design for treatment effect estimation
Understanding causality should be a core requirement of any attempt to build real impact through AI.
Representation Learning for High-Dimensional Data Collection under Local Differential Privacy
In this work, our contributions are two-fold: first, by adapting state-of-the-art techniques from representation learning, we introduce a novel approach to learning LDP mechanisms.
Explainability for fair machine learning
Moreover, motivated by the linearity of Shapley explainability, we propose a meta algorithm for applying existing training-time fairness interventions, wherein one trains a perturbation to the original model, rather than a new model entirely.
Human-interpretable model explainability on high-dimensional data
The importance of explainability in machine learning continues to grow, as both neural-network architectures and the data they model become increasingly complex.
Shapley explainability on the data manifold
Explainability in AI is crucial for model development, compliance with regulation, and providing operational nuance to predictions.
Asymmetric Shapley values: incorporating causal knowledge into model-agnostic explainability
We introduce a less restrictive framework, Asymmetric Shapley values (ASVs), which are rigorously founded on a set of axioms, applicable to any AI system, and flexible enough to incorporate any causal structure known to be respected by the data.
Binary JUNIPR: an interpretable probabilistic model for discrimination
We refer to this refined approach as Binary JUNIPR.
High Energy Physics - Phenomenology
Parenting: Safe Reinforcement Learning from Human Input
Autonomous agents trained via reinforcement learning present numerous safety concerns: reward hacking, negative side effects, and unsafe exploration, among others.
JUNIPR: a Framework for Unsupervised Machine Learning in Particle Physics
As a third application, JUNIPR models can reweight events from one (e. g. simulated) data set to agree with distributions from another (e. g. experimental) data set.
