Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
State Space Gaussian Processes with Non-Gaussian Likelihood
Hannes Nickisch, Arno Solin, Alexander Grigorevskiy
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 |
|---|---|---|---|
| Use state-space Kalman computations for GP inference instead of dense naive calculations | GPs with non-Gaussian likelihoods · Simulated likelihood/inference combinations (n=1000) | absolute error in posterior quantities: MAE typically equal up to 4--6 decimals | Improved |
| Interpolate transition and noise covariance matrices instead of evaluating every matrix exponential exactly | State-space GP · Simulated Matérn GP regression | computation time: factor of two | Improved |
| Increase interpolation grid size K (ablation) | State-space GP matrix exponential interpolation · Simulated Matérn GP regression | relative error in marginal likelihood | Improved |
| Use single-sweep ADF instead of full EP | State-space GP classification and Poisson regression · Simulated Poisson, logistic, and probit tasks | -log Z: 1200.11 vs 1206.59; 491.50 vs 525.46; 392.01 vs 433.75 | Mixed |
| Use a finer bin width for discretizing the event-time likelihood | Airline accident log Gaussian Cox process · Airline accidents | resolution of periodic intensity peaks | Improved |
Closest papers by meaning
- A general linear-time inference method for Gaussian Processes on one dimension (JMLR 2021 · similarity 0.92)
- Infinite-Horizon Gaussian Processes (NeurIPS 2018 · similarity 0.90)
- Computation-Aware Gaussian Processes: Model Selection And Linear-Time Inference (NeurIPS 2024 · similarity 0.87)
- Sparse Gaussian Processes: Structured Approximations and Power-EP Revisited (NeurIPS 2025 · similarity 0.87)
- Conditioning Gaussian Processes on Almost Anything (NeurIPS 2026 · similarity 0.87)
- Physics-Informed Variational State-Space Gaussian Processes (NeurIPS 2024 · similarity 0.86)
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.
