A Non–Convex Optimization Approach to Correlation Clustering

Authors: Erik Thiel, Morteza Haghir Chehreghani, Devdatt Dubhashi5159-5166

AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate the performance on both synthetic and real world data sets.
Researcher Affiliation Academia Chalmers University of Technology {erikthi, morteza.chehreghani, dubhashi}@chalmers.se
Pseudocode Yes Algorithm 1 Block-coordinate Frank-Wolfe Algorithm on a Product Domain
Open Source Code No The paper does not provide concrete access to source code for the methodology described in this paper.
Open Datasets Yes We demonstrate the performance on both synthetic and real world data sets. We formalize this idea using the World Color Survey dataset consisting of human labellings of 330 color tiles in different languages. In this section, we study the performance of different algorithms on several subsets of 20 newsgroup data collection.
Dataset Splits No The paper does not provide specific dataset split information for training, validation, or testing.
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 ancillary software details with version numbers needed to replicate the experiment.
Experiment Setup Yes The parameters are: n = 300, k = 10, p = 0.3, and the batch size b = 10 (i.e., q = 0.03). We fix the number of cluster k = 5. Then for a selected n, we assign each object randomly to one of the five different clusters. The weight between two nodes (objects) is a standard normal random variable.