Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Parametric Gaussian Process Regressors
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| directly optimize the predictive distribution with PPGPR | scalable GP regression · 12 UCI regression datasets | NLL: ~0.35 nats | Helped |
| use PPGPR | scalable GP regression · 12 UCI regression datasets | RMSE | Mixed |
| use SVGP instead of MAP | sparse GP regression · 12 UCI regression datasets | NLL: ~0.05 nats | Helped |
| increase variance-function capacity from delta to mean-field to full covariance (ablation) | PPGPR · UCI regression datasets | NLL | Helped |
| increase variance-function capacity from delta to mean-field to full covariance (ablation) | PPGPR · UCI regression datasets | RMSE | Hurt |
| decouple mean and variance functions and use separate inducing locations with diagonal covariance (ablation) | PPGPR-MFD · UCI regression datasets | RMSE | Helped |
| use the PPGPR objective instead of the corresponding inference objective (ablation) | PPGPR-delta versus MAP · UCI regression datasets | log likelihood | Helped |
| use the PPGPR objective instead of the corresponding inference objective (ablation) | PPGPR-Chol versus SVGP · UCI regression datasets | RMSE | Hurt |
| vary the regularization coefficient from 0 to 1 (ablation) | PPGPR · five smallest UCI datasets | RMSE: no more than 5% (relative) | No clear effect |
| vary the regularization coefficient from 0 to 1 (ablation) | PPGPR · five smallest UCI datasets | log likelihood: no more than 0.05 nats | No clear effect |
Closest papers by meaning
- Variational Gaussian Processes with Decoupled Conditionals (NeurIPS 2023 · similarity 0.91)
- Posterior and Computational Uncertainty in Gaussian Processes (NeurIPS 2022 · similarity 0.91)
- New Bounds for Sparse Variational Gaussian Processes (ICML 2025 · similarity 0.90)
- Robust and Conjugate Gaussian Process Regression (ICML 2024 · similarity 0.89)
- Turbocharging Gaussian Process Inference with Approximate Sketch-and-Project (NeurIPS 2025 · similarity 0.89)
- Amortized Structured Stochastic Variational Inference for Gaussian Process Latent Variable Models (NeurIPS 2026 · similarity 0.89)
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.