Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Thin and deep Gaussian processes
Daniel Augusto de Souza, Alexander V Nikitin, S. T. John, Magnus Ross, Mauricio A Álvarez, Marc Peter Deisenroth, João Paulo Pordeus Gomes, Diego Mesquita, César Lincoln Mattos
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 changed | Where | Measured | Effect |
|---|---|---|---|
| TDGP versus shallow SGP | TDGP · Synthetic 2D funnel | NLPD: -1.49 -> -3.57 | Improved |
| TDGP versus DNSGP | TDGP · Synthetic 2D funnel | MRAE: 0.12 -> 0.00 | Improved |
| TDGP versus DKL | TDGP · Synthetic 2D funnel | NLPD: 36.82 -> -3.57 | Improved |
| TDGP versus CDGP | TDGP · GEBCO bathymetry | NLPD: -0.44 -> -0.53 | Improved |
| TDGP versus DKL on MRAE | TDGP · GEBCO bathymetry | MRAE: 0.59 -> 0.66 | Hurt |
| TDGP versus SGP on MRAE | TDGP · GEBCO bathymetry | MRAE: 1.19 -> 0.66 | Improved |
| TDGP versus CDGP training time | TDGP · GEBCO bathymetry | training time (s): 570.8 -> 126.7 | Improved |
| TDGP versus shallow SGP training time | TDGP · GEBCO bathymetry | training time (s): 23.4 -> 126.7 | Hurt |
| TDGP on wine_red training time versus DNSGP | TDGP · UCI wine_red | training time (s): 650.8 -> 928.6 | Hurt |
| TDGP on UCI regression datasets versus other models | TDGP · UCI housing, concrete, wine-red, energy | NLPD | No clear effect |
| increase TDGP hidden width Q from 1 to D (ablation) | TDGP · UCI housing | training time | Hurt |
| increase TDGP width Q beyond 3 to 6 (ablation) | TDGP · UCI housing | test metrics | Mixed |
Closest papers by meaning
- Spherical Inducing Features for Orthogonally-Decoupled Gaussian Processes (ICML 2023 · similarity 0.87)
- Implicit Gaussian process representation of vector fields over arbitrary latent manifolds (ICLR 2024 · similarity 0.87)
- Non-Gaussian Gaussian Processes for Few-Shot Regression (NeurIPS 2021 · similarity 0.87)
- Residual Deep Gaussian Processes on Manifolds (ICLR 2025 · similarity 0.86)
- Computation-Aware Gaussian Processes: Model Selection And Linear-Time Inference (NeurIPS 2024 · similarity 0.86)
- Deep Random Features for Scalable Interpolation of Spatiotemporal Data (ICLR 2025 · similarity 0.85)
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.
