Research map / Probabilistic and scientific modeling / Causal discovery
Towards Scalable Bayesian Learning of Causal DAGs
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 a zeta transform for preprocessing candidate-parent score sums instead of the brute-force approach | Gadget DAG sampling · Preprocessing | time complexity: O(3^K K^2) -> O(2^K K) | Helped |
| Use the subset-sum data structure for preprocessing candidate-parent score sums | Gadget DAG sampling · Preprocessing | space complexity per node: O(3^K) -> O(2^K) | Helped |
| Increase the candidate-parent set size K (ablation) | Candidate-parent selection · Synthetic linear Gaussian DAGs and UCI discrete data | coverage | Helped |
| Increase the sample size from N=50 to N=200 (ablation) | Optimal candidate-parent selection · Synthetic linear Gaussian DAGs, K=12 | coverage: 0.90 at K = 12 and N = 200 | Helped |
| Use Greedy candidate-parent selection rather than the other tested heuristics | Candidate-parent selection · Synthetic linear Gaussian DAGs | coverage | Helped |
| Use Greedy or Back&Forth rather than the other tested heuristics | Candidate-parent selection · UCI discrete data | coverage | Helped |
| Use Gadget rather than the compared MCMC methods or exact algorithm | Ancestor-relation inference · Binary Bayesian networks, 20 and 50 nodes | area under the ROC curve | Mixed |
| Use Beeps with Gadget for causal-effect estimation | Linear Gaussian Bayesian networks · Marginal causal effects | mean squared error | Helped |
| Use Beeps rather than the available IDA-based methods | Joint causal-effect estimation · Linear Gaussian Bayesian networks | mean squared error | Helped |
| Use Beeps with Gadget rather than the compared methods | Causal-effect estimation · ARTH150, 107 nodes | mean squared error | Helped |
Closest papers by meaning
- Causal Discovery via Bayesian Optimization (ICLR 2025 · similarity 0.92)
- Sample Efficient Bayesian Learning of Causal Graphs from Interventions (NeurIPS 2024 · similarity 0.92)
- Estimating Interventional Distributions with Uncertain Causal Graphs through Meta-Learning (NeurIPS 2025 · similarity 0.89)
- Coarsening Linear Non-Gaussian Causal Models with Cycles (NeurIPS 2026 · similarity 0.89)
- Interventional Causal Discovery in a Mixture of DAGs (NeurIPS 2024 · similarity 0.89)
- A Meta-Learning Approach to Bayesian Causal Discovery (ICLR 2025 · similarity 0.89)
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.