Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Bayesian Spiked Laplacian Graphs
Leo L Duan, George Michailidis, Mingzhou Ding
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 |
|---|---|---|---|
| use spiked Laplacian instead of direct decomposition of observed Laplacian | spectral community detection · single graph, spectral gap 0.05 | NMI: 0.42 -> 0.58 | Improved |
| use spiked Laplacian instead of direct decomposition of observed Laplacian | spectral community detection · single graph, spectral gap 0.1 | NMI: 0.78 -> 0.88 | Improved |
| use spiked Laplacian instead of direct decomposition of observed Laplacian | spectral community detection · single graph, spectral gap 0.01 | NMI: 0.36 -> 0.40 | Improved |
| use spiked Laplacian model instead of stochastic block model | community detection · n=100 | NMI: 0.65 -> 0.84 | Improved |
| use spiked Laplacian model instead of Bayesian SBM | community detection · n=100 | NMI: 0.70 -> 0.84 | Improved |
| use spiked Laplacian model instead of SBM | community detection · n=500 | NMI: 0.87 -> 0.86 | No clear effect |
| use spiked Laplacian model instead of SBM | community detection · n=300 | NMI: 0.84 -> 0.90 | Improved |
| use spiked Laplacian graphs instead of averaging graphs then SBM | heterogeneous graph collection · 500 simulated graphs | NMI: 0.21 -> 0.85 | Improved |
| use spiked Laplacian graphs instead of clustering graphs then SBMs | heterogeneous graph collection · 500 simulated graphs | NMI: 0.67 -> 0.85 | Improved |
| use spiked Laplacian graphs instead of individual SBMs | heterogeneous graph collection · 500 simulated graphs | NMI: 0.45 -> 0.85 | Improved |
| use individual SBMs instead of spiked Laplacian graphs | heterogeneous graph collection · 500 simulated graphs | RMSE (x10^-3): 1.9 -> 1.2 | Improved |
| over-specify T (T=10 or T=30 versus true number of eigenvalue components) (ablation) | spiked Laplacian model · single graph, three communities | first eigenvalue estimates | No clear effect |
Closest papers by meaning
- Bayesian Spanning Tree: Estimating the Backbone of the Dependence Graph (JMLR 2023 · similarity 0.89)
- Coupled Harmonic Bases for Longitudinal Characterization of Brain Networks (CVPR 2016 · similarity 0.87)
- High-Dimensional Inference for Cluster-Based Graphical Models (JMLR 2020 · similarity 0.87)
- Recovering shared structure from multiple networks with unknown edge distributions (JMLR 2022 · similarity 0.87)
- Efficient Variational Bayes Learning of Graphical Models With Smooth Structural Changes (TPAMI 2022 · similarity 0.87)
- Nonlinear Laplacians: Tunable principal component analysis under directional prior information (NeurIPS 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.
