AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

Computationally and statistically efficient learning of causal Bayes nets using path queries

Kevin Bello, Jean Honorio

NeurIPS 2018 · Probabilistic and scientific modeling › Causal discovery · approach: Constraint-based structure learning with latents · 2 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use noisy interventional path queries with single-vertex interventions instead of large multiple-vertex interventionsLearning the transitive reduction of general discrete DAGs · General DAGs, discretenumber of experiments: $\bO{2^n \log n}$ -> $\bO{n^2 2^t}$Improved
Use the rooted-tree query strategy rather than the general-DAG query strategyLearning rooted trees · Rooted treesnumber of path queries: $n^2$ -> $\bO{n \log n}$Improved
Account for imperfect discrete interventions using conditional samplesNoisy path-query algorithm for discrete CBNs · Discrete CBNs under imperfect interventionssample complexity: $\bO{\frac{1}{\gamma^2} \left( \ln n + \ln \frac{ r}{\delta} \right)}$ -> $\bO{\frac{1}{\alpha\gamma^2} \left( \ln n + \Hurt
Use imperfect rather than perfect interventions in continuous CBNsNoisy path-query algorithm for continuous CBNs · Continuous CBNs under imperfect interventionssample complexity dependence on varianceNo clear effect

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy