Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Variational Learning on Aggregate Outputs with Gaussian Processes
Ho Chung Law, Dino Sejdinovic, Ewan Cameron, Tim Lucas, Seth Flaxman, Katherine Battle, Kenji Fukumizu
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 VBAgg instead of Nyström or a neural network | Poisson regression on aggregate outputs · Swiss roll, varying number of training bags | individual NLL | Improved |
| Use VBAgg instead of Nyström or a neural network | Poisson regression on aggregate outputs · Swiss roll, varying number of individuals per bag | individual NLL | Improved |
| Increase the number of individuals per bag | VBAgg · Swiss roll, varying number of individuals per bag | individual NLL | No clear effect |
| Aggregate individual covariates into bag-level covariates before training | bag-pixel model · Swiss roll, varying number of training bags | individual NLL | Hurt |
| Use Nyström instead of VBAgg | Poisson regression on aggregate outputs · Malaria incidence, test bags | bag NLL | No clear effect |
| Use VBAgg or Nyström instead of a neural network | Poisson regression on aggregate outputs · Malaria incidence, test bags | bag NLL | Improved |
| Use the squared link (VBAgg-Sq) instead of the exponential link (VBAgg-Exp) (ablation) | VBAgg · Swiss roll | calibration | Improved |
| Use the exponential link (VBAgg-Exp) instead of the squared link (VBAgg-Sq) (ablation) | VBAgg · Swiss roll | individual NLL | No clear effect |
| Tune VBAgg parameters using the training ELBO rather than validation bag NLL (ablation) | VBAgg · Toy experiments | prediction performance | No clear effect |
| Use the squared link (VBAgg-Sq) instead of the exponential link (VBAgg-Exp) (ablation) | VBAgg · Malaria incidence, pixel-level predictions | spatial pattern quality | Improved |
Closest papers by meaning
- Robust and Scalable Gaussian Process Regression and Its Applications (CVPR 2023 · similarity 0.86)
- Multi-task Learning for Aggregated Data using Gaussian Processes (NeurIPS 2019 · similarity 0.86)
- Environment-Robust Representation Learning with Empirical Bayes (NeurIPS 2026 · similarity 0.84)
- Minimal Variance Model Aggregation: A principled, non-intrusive, and versatile integration of black box models (ICLR 2025 · similarity 0.84)
- Amortized Structured Stochastic Variational Inference for Gaussian Process Latent Variable Models (NeurIPS 2026 · similarity 0.84)
- Sparse Variational Student-t Processes (AAAI 2024 · similarity 0.83)
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.
