Research map / Probabilistic and scientific modeling / Molecules and proteins
Interaction Measures, Partition Lattices and Kernel Tests for High-Order Interactions
Zhaolu Liu, Robert Peach, Pedro A. M. Mediano, Mauricio Barahona
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 |
|---|---|---|---|
| eliminate partitions with singletons via kernel centring (ablation) | Streitberg estimator · d=8 variables | number of terms: 4140 -> 715 | Improved |
| eliminate partitions with singletons via kernel centring (ablation) | Streitberg estimator · d=4 variables | number of terms: 15 -> 4 | Improved |
| centre kernel matrices for Lancaster estimator (ablation) | Lancaster estimator · d=8 variables | number of terms: 248 -> 1 | Improved |
| use Streitberg instead of dHSIC | kernel interaction test · 5-way XOR, joint independence, n=80 | power | Improved |
| increase sample size for dHSIC from n=80 to n=1000 | dHSIC · 5-way XOR, joint independence | power: 80 -> 1000 | Improved |
| use Streitberg instead of dHSIC for factorisation without singletons P12 P345 | Streitberg test · 5-variable Gaussian | detection of factorisation | Improved |
| use Lancaster instead of Streitberg for factorisation without singletons P12 P345 | Lancaster test · 5-variable Gaussian | detection of factorisation | Hurt |
| use modified dHSIC for sub-hypotheses instead of Streitberg | modified dHSIC · 5-way XOR sub-hypotheses | power | Hurt |
| reduce sample size | dHSIC vs Lancaster/Streitberg · XOR example | null rejection rate | Hurt |
| use Streitberg rather than Lancaster number of sub-hypotheses | interaction tests · d=4 | number of sub-hypotheses: 4 -> 7 | Hurt |
| ground truth P1 P2 P34 instead of P1 P234 for partial rejections | Streitberg sub-tests · d=4 | number of rejections: 6 -> 4 | Mixed |
| regions within same resting state network instead of random | Streitberg test · HCP fMRI | fraction of significant interactions | Improved |
Closest papers by meaning
- Permutation-Free High-Order Interaction Tests (ICML 2025 · similarity 0.90)
- The Kernel Interaction Trick: Fast Bayesian Discovery of Pairwise Interactions in High Dimensions (ICML 2019 · similarity 0.89)
- The SKIM-FA Kernel: High-Dimensional Variable Selection and Nonlinear Interaction Discovery in Linear Time (JMLR 2023 · similarity 0.87)
- Fast and More Powerful Selective Inference for Sparse High-Order Interaction Model (AAAI 2022 · similarity 0.86)
- S$\Omega$I: Score-based O-INFORMATION Estimation (ICML 2024 · similarity 0.84)
- Benchmarking Structural Inference Methods for Interacting Dynamical Systems with Synthetic Data (NeurIPS 2024 · similarity 0.84)
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.
