Research map / Learning theory and algorithms / Matrix and tensor methods
Low-Rank Factorization of Determinantal 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 low-rank kernel factorization (K=30) instead of full-rank kernel | DPP for basket completion · Amazon Baby Registry (15 categories) | average test log-likelihood: full-rank better in 9 categories, low-rank in 6 | No clear effect |
| use low-rank kernel factorization (K=15) | DPP for basket completion · MS Store | average test log-likelihood: -15.10 -> -15.23 | Hurt |
| use low-rank factorization to train DPP | DPP kernel learning · Amazon Baby Registry | training time: 8.9x faster on average | Helped |
| use low-rank factorization for conditional prediction | DPP basket completion · Amazon Baby Registry | prediction time per basket: 2.55 ms -> 0.39 ms | Helped |
| use low-rank factorization for conditional prediction on larger catalog | DPP basket completion · MS Store (2097 items) | prediction time per basket: 1.66 s -> 83.6 ms (19.9x) | Helped |
| increase number of latent trait dimensions K up to 14 and beyond (ablation) | low-rank DPP · Amazon apparel | average test log-likelihood | Mixed |
| low-rank V matrix instead of full kernel storage | DPP model · MS Store | memory: 35.18 MB -> 251.6 KB | Helped |
Closest papers by meaning
- Learning Nonsymmetric Determinantal Point Processes (NeurIPS 2019 · similarity 0.92)
- Dynamic Determinantal Point Processes (AAAI 2018 · similarity 0.90)
- Kronecker Determinantal Point Processes (NeurIPS 2016 · similarity 0.90)
- State-of-art minibatches via novel DPP kernels: discretization, wavelets, and rough objectives (NeurIPS 2026 · similarity 0.87)
- Small coresets via negative dependence: DPPs, linear statistics, and concentration (NeurIPS 2024 · similarity 0.85)
- Recommendations with Sparse Comparison Data: Provably Fast Convergence for Nonconvex Matrix Factorization (ICML 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.