Research map / Learning theory and algorithms / Prediction and decision losses
Bayesian Batch Active Learning as Sparse Subset Approximation
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 |
|---|---|---|---|
| use ACS-FW batch construction instead of random selection | neural linear model · yacht | test RMSE: 1.272 -> 1.031 | Helped |
| use ACS-FW instead of naive top-b MAXENT batches | neural linear model · energy | test RMSE: 1.443 -> 0.855 | Helped |
| use ACS-FW instead of naive top-b MAXENT batches | neural linear model · yacht | test RMSE: 0.923 -> 1.031 | Hurt |
| use ACS-FW instead of naive top-b MAXENT batches | neural linear model · year | test RMSE: 13.030 -> 12.194 | Helped |
| use sequential greedy MAXENT-SG instead of ACS-FW batches | neural linear model · boston | test RMSE: 3.799 -> 3.458 | Helped |
| use ACS-FW instead of RANDOM on large classification tasks | ResNet18 neural linear model · cifar10, SVHN, Fashion MNIST | accuracy | Helped |
| use ACS-FW instead of K-Center/BALD/MaxEnt | ResNet18 neural linear model · cifar10, SVHN, Fashion MNIST | accuracy | No clear effect |
| use ACS-FW instead of MAXENT on batch construction runtime | neural linear model · power | total runtime (s): 617.0 -> 179.8 | Helped |
Closest papers by meaning
- Batch Active Learning at Scale (NeurIPS 2021 · similarity 0.93)
- Scalable Batch-Mode Deep Bayesian Active Learning via Equivalence Class Annealing (ICLR 2023 · similarity 0.91)
- BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning (NeurIPS 2019 · similarity 0.90)
- Active Statistical Inference (ICML 2024 · similarity 0.86)
- Active Learning Guided by Efficient Surrogate Learners (AAAI 2024 · similarity 0.85)
- Transductive Active Learning: Theory and Applications (NeurIPS 2024 · 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.