Research map / 3D vision / Human pose estimation
RatioSketch: Towards More Accurate Frequency Estimation in Data Streams via a Lightweight Neural Network
Mengbo Wang, Zhuochen Fan, Dayu Wang, Guorui Xie, Qing Li, Zeyu Luan, Yong Jiang, Tong Yang, Mingwei Xu
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 |
|---|---|---|---|
| add RatioSketch correction module to CMS | CMS · CAIDA, 5MB, all items | AAE: 31.0x lower AAE | Improved |
| add RatioSketch correction module to TowerSketch | TowerSketch · CAIDA, 5MB, all items | ARE: 172.4x lower ARE | Improved |
| use RatioSketch-corrected sketch instead of LegoSketch | LegoSketch · CAIDA, 5MB, all items | ARE: 2301.5x lower ARE (TowerSketch+RatioSketch) | Improved |
| add RatioSketch correction for frequent items (top 20%) | CMS · CAIDA, 5MB, frequent items | AAE: 28.0x lower AAE | Improved |
| advantage larger at small memory budget | RatioSketch · CAIDA and others | AAE/ARE | Improved |
| add RatioSketch correction | CMS and TowerSketch · CAIDA | query throughput | Hurt |
| add RatioSketch correction | CMS and TowerSketch · CAIDA | store throughput | No clear effect |
| add global information branch (ablation) | RatioSketch · CAIDA, DC, Webdocs | AAE/ARE | Improved |
| add local information branch (ablation) | RatioSketch · CAIDA, DC, Webdocs | AAE/ARE | Improved |
| add structural (load) information branch (ablation) | RatioSketch · CAIDA, DC, Webdocs | AAE/ARE | Improved |
| use 64 counters per row in global encoder instead of fewer or more (ablation) | RatioSketch Global Encoder · design choice | accuracy | Mixed |
Closest papers by meaning
- Lego Sketch: A Scalable Memory-augmented Neural Network for Sketching Data Streams (ICML 2025 · similarity 0.85)
- Learning-Augmented Frequent Directions (ICLR 2025 · similarity 0.84)
- CTSketch: Compositional Tensor Sketching for Scalable Neurosymbolic Learning (NeurIPS 2025 · similarity 0.82)
- Slower Generalization, Faster Memorization: A Sweet Spot in Algorithmic Learning (NeurIPS 2026 · similarity 0.82)
- Sketchy Moment Matching: Toward Fast and Provable Data Selection for Finetuning (NeurIPS 2024 · similarity 0.82)
- Diversity-Driven Synthesis: Enhancing Dataset Distillation through Directed Weight Adjustment (NeurIPS 2024 · similarity 0.82)
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.
