Understanding Regularized Spectral Clustering via Graph Conductance

Authors: Yilin Zhang, Karl Rohe

NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We provide simulations and data examples to illustrate these results.
Researcher Affiliation Academia Yilin Zhang Department of Statistics University of Wisconsin-Madison Madison, WI 53706 yilin.zhang@wisc.edu Karl Rohe Department of Statistics University of Wisconsin-Madison Madison, WI 53706 karl.rohe@wisc.edu
Pseudocode No The paper does not contain any pseudocode or clearly labeled algorithm blocks.
Open Source Code No The paper does not provide any explicit statements about releasing source code for the methodology described, nor does it include a direct link to a code repository.
Open Datasets Yes This section studies 37 example networks from http://snap.stanford.edu/data [14]. [14] Jure Leskovec and Andrej Krevl. SNAP Datasets: Stanford large network dataset collection. http://snap.stanford.edu/data, June 2014.
Dataset Splits No In this simulation, half of the edges are removed from the graph and placed into a testing-set . Refer to the remaining edges as the training-edges . While a train/test split is described, there is no explicit mention of a validation set or split.
Hardware Specification No The paper does not provide any specific details about the hardware (e.g., GPU/CPU models, memory) used for running the experiments.
Software Dependencies Yes All eigen-computations are performed with rARPACK [13, 18]. ... [18] Yixuan Qiu, Jiali Mei, and authors of the ARPACK library. See file AUTHORS for details. r ARPACK: Solvers for Large Scale Eigenvalue and SVD Problems, 2016. R package version 0.11-0.
Experiment Setup Yes Throughout all simulations, the regularization parameter τ is set to be the average degree of the graph. This is not optimized, but is instead a previously proposed heuristic [17]. As defined in Section 2 Equation 2.1, the partitions are constructed by scanning through the second eigenvector. ... Running times are from rARPACK computing two eigenvectors of D 1/2 τ AD 1/2 τ and D 1/2AD 1/2 using the default settings.