AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Gaussian Processes with Linear Operator Inequality Constraints

Christian Agrell

JMLR 2019 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: PAC-Bayes and decoupled GP learning · 39 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add linear inequality constraints to the GPunconstrained GP · Robot arm functionQ2: 0.7558 -> 0.8842Improved
add linear inequality constraints to the GPunconstrained GP · Robot arm functionPVA: 3.03 -> 2.85Improved
add linear inequality constraints to the GPunconstrained GP · Robot arm functionAWoCI: 0.99 -> 0.54Improved
use moment approximation for inference as well as virtual location search (Moment approx. 2) instead of sampling inference (ablation)constrained GP · Robot arm functionAWoCI: 0.54 -> 0.83Hurt
use moment approximation only for virtual location search (Moment approx. 1) (ablation)constrained GP · Robot arm functionQ2: 0.8842 -> 0.8844No clear effect
use correlation-free approximation for virtual location search (ablation)constrained GP · Robot arm functionCPU time for 10^4 samples (s): 24.8 -> 21.1Improved
use correlation-free approximation for virtual location search (ablation)constrained GP · Robot arm functionQ2: 0.8842 -> 0.8775No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy