Graph Clustering: Block-models and model free results
Authors: Yali Wan, Marina Meila
NeurIPS 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | 6 Experimental evaluation Given G, we obtain a clustering C0 by spectral clustering [15]. Then we calculate clustering C by perturbing C0 with gradually increasing noise. For each C, we construct PFM (C, G)and SBM(C, G) model, and further compute , δ and δ0. If δ δ0, C is guaranteed to be stable by the theorems. In the remainder of this section, we describe the data generating process for the simulated datasets and the results we obtained. |
| Researcher Affiliation | Academia | Yali Wan Department of Statistics University of Washington Seattle, WA 98195-4322, USA yaliwan@washington.edu Marina Meil a Department of Statistics University of Washington Seattle, WA 98195-4322, USA mmp@stat.washington.edu |
| Pseudocode | Yes | PFM Estimation Algorithm Input Graph G with ˆA, ˆD, ˆL, ˆY , ˆΛ, clustering C with indicator matrix Z. Output (A, L) = PFM(G, C) 1. Construct an orthogonal matrix derived from Z. YZ = ˆD1/2ZC 1/2, with C = ZT ˆDZ the column normalization of Z. (5) 2. Project YZ on ˆY and perform Singular Value Decomposition. F = Y T Z ˆY = UΣV T (6) 3. Change basis in R(YZ) to align with ˆY . Y = YZUV T . Complete Y to an orthonormal basis [Y B] of Rn. (7) 4. Construct Laplacian L and edge probability matrix A. L = Y ˆΛY T + (BBT )ˆL(BBT ), A = ˆD1/2L ˆD1/2. (8) |
| Open Source Code | No | The paper does not provide explicit statements or links for open-source code for the described methodology. |
| Open Datasets | Yes | Political Blogs Dataset A directed network A of hyperlinks between weblogs on US politics, compiled from online directories by Adamic and Glance [2], where each blog is assigned a political leaning, liberal or conservative, based on its blog content. The network A contains 1490 blogs. |
| Dataset Splits | No | The paper describes dataset generation parameters and cluster sizes, but does not provide specific training, validation, or test dataset splits. |
| Hardware Specification | No | The paper does not provide specific hardware details used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers. |
| Experiment Setup | No | The 'Experiment Setup' section describes the data generation process and computed quantities (ε, δ, δ0) but does not provide specific hyperparameter values or detailed training configurations for any model. |