Research map / Label-efficient and robust learning / Recognition and long-tail
Multi-Level Deep Cascade Trees for Conversion Rate Prediction in Recommendation System
Hong Wen, Jing Zhang, Quan Lin, Keping Yang, Pipei Huang
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 |
|---|---|---|---|
| Use level-by-level cascade learning with cross-entropy leaf features instead of a single GBDT | CVR prediction · Offline test set | AUC: 0.7692 -> 0.7942 | Improved |
| Add parallel diverse ldcTree ensembles | ldcTree · Offline test set | AUC: 0.7942 -> 0.8121 | Improved |
| Add staged use of weak- and strong-correlation features | EldcTree · Offline test set | AUC: 0.8121 -> 0.8315 | Improved |
| Increase tree depth from 4 to 8 (ablation) | ldcTree · Validation set | AUC: 0.783 -> 0.794 | Improved |
| Increase tree depth from 6 to 8 (ablation) | ldcTree · Validation set | AUC: 0.789 -> 0.794 | Improved |
| Increase tree depth from 8 to 10 (ablation) | ldcTree · Validation set | AUC: 0.794 -> 0.797 | Improved |
| Increase tree depth from 8 to 10 (ablation) | ldcTree · Validation set | Time Cost(s): 15 -> 26 | Hurt |
| Use ldcTree instead of Naive GBDT | Online CVR prediction · Online recommendation system, four days | CVR: 104.3%, 104.1%, 103.8%, 103.9% | Improved |
| Add ensemble structure to ldcTree | Online CVR prediction · Online recommendation system, four days | CVR: 107.1%, 107.4%, 106.3%, 106.8% | Improved |
| Use F-EldcTree in place of the baseline model | Online CVR prediction · Online recommendation system | CVR: 12 percent | Improved |
Closest papers by meaning
- Learning Deep Tree-Based Retriever for Efficient Recommendation: Theory and Method (TPAMI 2025 · similarity 0.87)
- Joint Optimization of Tree-based Index and Deep Model for Recommender Systems (NeurIPS 2019 · similarity 0.84)
- Recommender Forest for Efficient Retrieval (NeurIPS 2022 · similarity 0.83)
- A Hybrid Collaborative Filtering Model with Deep Structure for Recommender Systems (AAAI 2017 · similarity 0.82)
- Learning Cascade Ranking as One Network (ICML 2025 · similarity 0.81)
- Hgformer: Hyperbolic Graph Transformer for Collaborative Filtering (ICML 2025 · similarity 0.81)
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.
