Research map / Optimization / Optimizers
Active Mini-Batch Sampling Using Repulsive Point Processes
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 repulsive Poisson Disk sampling instead of standard random mini-batch sampling | SGD · MNIST | test error rate | Helped |
| Use repulsive Poisson Disk sampling instead of standard random mini-batch sampling | SGD · Speech Command Recognition | validation accuracy | Helped |
| Use Vanilla PDS instead of standard SGD | SGD · Oxford Flower | test accuracy | Helped |
| Use Poisson Disk sampling instead of k-DPP | Oxford Flower mini-batch sampling · Oxford Flower | sampling time (sec): 7.1680 -> 0.0461; 29.687 -> 0.0795; 58.4086 -> 0.1048; 189.0303 -> 0.1657; 436.4746 -> 0.2391 | Helped |
| Use Poisson Disk sampling instead of Fast k-DPP | Oxford Flower mini-batch sampling · Oxford Flower | sampling time (sec): 0.1032 -> 0.0461; 0.3312 -> 0.0795; 0.6512 -> 0.1048; 1.8745 -> 0.1657; 4.1964 -> 0.2391 | Helped |
| Add adaptive repulsion that exempts points near decision boundaries (ablation) | Vanilla PDS · MNIST | test error rate | No clear effect |
| Bias sampling toward points with higher mingling indices (ablation) | Vanilla PDS (Dense PDS) · MNIST | test error rate | Helped |
| Anneal sampling density from easy examples toward higher-mingling examples (ablation) | Dense PDS (Anneal PDS) · MNIST | test error rate | Helped |
| Add adaptive repulsion that exempts points with nonzero mingling indices (ablation) | Vanilla PDS (Easy PDS) · Speech Command Recognition | validation accuracy | No clear effect |
| Bias sampling toward points with higher mingling indices (ablation) | Vanilla PDS (Dense PDS) · Speech Command Recognition | validation accuracy | No clear effect |
Closest papers by meaning
- Determinantal point processes based on orthogonal polynomials for sampling minibatches in SGD (NeurIPS 2021 · similarity 0.92)
- State-of-art minibatches via novel DPP kernels: discretization, wavelets, and rough objectives (NeurIPS 2026 · similarity 0.88)
- Stochastic Optimization with Arbitrary Recurrent Data Sampling (ICML 2024 · similarity 0.88)
- The Sample Complexity of Gradient Descent in Stochastic Convex Optimization (NeurIPS 2024 · similarity 0.88)
- Demystifying SGD with Doubly Stochastic Gradients (ICML 2024 · similarity 0.87)
- Faster Sampling via Stochastic Gradient Proximal Sampler (ICML 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.