Research map / Optimization / Optimizers
StochasticRank: Global Optimization of Scale-Free Discrete Functions
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 StochasticRank instead of LambdaMART | LTR on synthetic data · Synthetic data | NDCG@3: 0.903 -> 0.917 | Helped |
| Use StochasticRank instead of the strongest baseline | LTR · Yahoo Set 1 | NDCG@5: 74.75 -> 74.92 | Helped |
| Use StochasticRank instead of the strongest baseline | LTR · Yahoo Set 1 | MRR: 90.30 -> 90.97 | Helped |
| Use StochasticRank instead of the strongest baseline | LTR · Yahoo Set 2 | NDCG@5: 73.91 -> 74.15 | Helped |
| Use StochasticRank instead of the strongest baseline | LTR · Yahoo Set 2 | MRR: 92.16 -> 93.56 | Helped |
| Use StochasticRank instead of the strongest baseline | LTR · WEB10K | NDCG@5: 48.47 -> 48.53 | Helped |
| Use StochasticRank instead of the strongest reported alternative | LTR · WEB10K | MRR: 83.08 -> 83.30 | Helped |
| Use SR-R1soft instead of the strongest baseline | LTR · WEB30K | NDCG@5: 49.52 -> 49.67 | Helped |
| Use SR-R1soft instead of the strongest baseline | LTR · WEB30K | MRR: 83.79 -> 85.19 | Helped |
| Replace REINFORCE with CCS (ablation) | StochasticRank · Yahoo Set 1 | NDCG@5: 70.74 -> 71.89 | Helped |
| Add scale-free acceleration to CCS (ablation) | StochasticRank · Yahoo Set 1 | NDCG@5: 71.89 -> 74.55 | Helped |
| Add SGLB to CCS+SFA (ablation) | StochasticRank · Yahoo Set 1 | NDCG@5: 74.55 -> 74.68 | Helped |
Closest papers by meaning
- PiRank: Scalable Learning To Rank via Differentiable Sorting (NeurIPS 2021 · similarity 0.87)
- Stochastic Continuous Submodular Maximization: Boosting via Non-oblivious Function (ICML 2022 · similarity 0.87)
- Which Tricks are Important for Learning to Rank? (ICML 2023 · similarity 0.86)
- SGLB: Stochastic Gradient Langevin Boosting (ICML 2021 · similarity 0.86)
- Tuning-Free Stochastic Optimization (ICML 2024 · similarity 0.85)
- Stochastic Gradients under Nuisances (NeurIPS 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.