Research map / Probabilistic and scientific modeling / Causal discovery
Causal Structure Learning for Sparse Matrix Fill-in Reduction
Ziwei Li, Shuzi Niu, Tao Yuan, Huiyuan Li
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 causal triplet loss lCTS instead of PFM loss (ablation) | MgKGAT · CPS+CFD ablation | fill-in ratio: 119.02 -> 107.51 | Improved |
| add multigrid architecture to KGAT (ablation) | KGAT · CPS ablation | fill-in ratio: 167.41 -> 160.43 | Improved |
| add multigrid architecture to KGAT (ablation) | KGAT · CFD ablation | fill-in ratio: 52.83 -> 54.59 | Hurt |
| add multigrid architecture to GUnet-style network (ablation) | GUnet vs MgKGAT · CPS+CFD ablation | fill-in ratio: 164.19 -> 107.51 | Improved |
| add vertex embedding network on top of spectral embedding (ablation) | SE · CPS+CFD ablation | fill-in ratio: 118.74 -> 107.51 | Improved |
| use KAN-based KGAT instead of GAT in multigrid (ablation) | MgGAT · CPS+CFD ablation | fill-in ratio: 109.37 -> 107.51 | Improved |
| use KGAT instead of SAGE in multigrid (ablation) | MgSAGE · CPS+CFD ablation | fill-in ratio: 111.28 -> 107.51 | Improved |
| CTS versus PFM on LU time | CTS · CFD | LU factorization time (s): 7.52 -> 5.02 | Improved |
| CTS versus UDNO on SP | CTS · SP | fill-in ratio: 52.17 vs 48.64 | Hurt |
Closest papers by meaning
- Compress to Impress: Efficient LLM Adaptation Using a Single Gradient Step on 100 Samples (NeurIPS 2025 · similarity 0.84)
- Schwarz–Schur Involution: Lightspeed Differentiable Sparse Linear Solvers (ICML 2025 · similarity 0.84)
- Two Sparse Matrices are Better than One: Sparsifying Neural Networks with Double Sparse Factorization (ICLR 2025 · similarity 0.84)
- Connectivity Shapes Implicit Regularization in Matrix Factorization Models for Matrix Completion (NeurIPS 2024 · similarity 0.84)
- DCILP: A Distributed Approach for Large-Scale Causal Structure Learning (AAAI 2025 · similarity 0.84)
- Be a Goldfish: Forgetting Bad Conditioning in Sparse Linear Regression via Variational Autoencoders (ICML 2025 · similarity 0.83)
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.
