Research map / Label-efficient and robust learning / Recognition and long-tail
CAFE: Adaptive VDI Workload Prediction with Multi-Grained Features
Yao Zhang, Wen-Ping Fan, Xuan Wu, Hua Chen, Bin-Yang Li, Min-Ling Zhang
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 multi-grained features (CAFE vs single-granularity NF-GBDT) (ablation) | CAFE · Venus-D1, dt=30 | NMAE: 0.0969 -> 0.0625 | Improved |
| add multi-grained features (CAFE vs NF-GBDT) (ablation) | CAFE · Uranus-D1, dt=30 | NMAE: 0.2595 -> 0.1224 | Improved |
| add multi-grained features (CAFE vs NF-GBDT) (ablation) | CAFE · Mars-D2, dt=30 | NMAE: 0.3081 -> 0.1260 | Improved |
| add multi-grained features (CAFE vs NF-GBDT) (ablation) | CAFE · Pluto-D2, dt=30 | NMAE: 0.0640 -> 0.0775 | Hurt |
| add multi-grained features (CAFE vs NF-GBDT) (ablation) | CAFE · Uranus-D1, dt=60 | NMAE: 0.3321 -> 0.1639 | Improved |
| add multi-grained features (CAFE vs NF-GBDT) (ablation) | CAFE · Mars-D2, dt=60 | NRMSE: 0.3721 -> 0.1583 | Improved |
| CAFE vs Holt-Winters (ablation) | CAFE · Mars-D2, dt=60 | NMAE: 1.8100 -> 0.1550 | Improved |
| CAFE vs Prophet (ablation) | CAFE · Venus-D1, dt=30 | NMAE: 0.2981 -> 0.0625 | Improved |
| longer prediction span dt=60 instead of dt=30 | CAFE · Venus-D2 | NMAE: 0.0953 -> 0.1589 | Hurt |
| CAFE ranks first over all configurations | CAFE · 16 data sets x 4 metrics | rank: 82.8% | Improved |
| CAFE vs others, shorter span | CAFE · dt=30, 8 data sets x 4 metrics | rank 1st share: 87.5% | Improved |
| CAFE vs others, longer span | CAFE · dt=60, 8 data sets x 4 metrics | rank 1st share: 78.1% | Improved |
Later papers that compared with this method
2 later accepted papers report CAFE as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Bites of Tomorrow: Personalized Recommendations for a Healthier and Greener Plate (AAAI 2025 · better on 1 of 4 shared results)
- Improved Distribution Matching for Dataset Condensation (CVPR 2023 · better on 1 of 1 shared result)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Optimizing IT FinOps and Sustainability through Unsupervised Workload Characterization (AAAI 2024 · similarity 0.84)
- DCcluster-Opt: Benchmarking Dynamic Multi-Objective Optimization for Geo-Distributed Data Center Workloads (NeurIPS 2025 · similarity 0.78)
- Spatial-Temporal Heterogenous Graph Contrastive Learning for Microservice Workload Prediction (AAAI 2025 · similarity 0.78)
- Tackling the Data-Parallel Load Balancing Bottleneck in LLM Serving: Practical Online Routing at Scale (NeurIPS 2026 · similarity 0.78)
- MAPS: Memory-Aware Predictive Scheduling Framework for Large Language Model Serving (ICML 2026 · similarity 0.77)
- FastPERT: Towards Fast Microservice Application Latency Prediction via Structural Inductive Bias over PERT Networks (AAAI 2025 · similarity 0.77)
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.
