Local Correlation Clustering with Asymmetric Classification Errors

Authors: Jafar Jafarov, Sanchit Kalhan, Konstantin Makarychev, Yury Makarychev

ICML 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical In this paper we study the task of minimizing local objectives (Definition 1.1) under the Correlation Clustering with Asymmetric Classification Errors model (Definition 1.2). Our main result is an O (1/α) 1/2 1/2p log 1/α approximation algorithm for minimizing the ℓp norm of the disagreements vector, which we now state. Theorem 1.3. There exists a polynomial-time O (1/α) 1/2 1/2p log 1/α -approximation algorithm for the ℓp objective in the Correlation Clustering with Asymmetric Classification Errors model. ...We compliment our positive result (Theorem 1.3) by showing that it is likely to be the best possible based on the natural convex program, by providing an almost matching integrality gap.
Researcher Affiliation Academia 1University of Chicago, Chicago, IL, USA 2Northwestern University, Evanston, IL, USA 3TTIC, Chicago, IL, USA.
Pseudocode Yes Algorithm 1 Correlation Clustering Algorithm ... Algorithm 2 Metric Space Partitioning Scheme ... Algorithm 3 Cluster Select
Open Source Code No The paper does not provide any statements about open-source code availability or links to repositories.
Open Datasets No The paper is theoretical and does not involve empirical evaluation on datasets. Therefore, no information regarding publicly available datasets is provided.
Dataset Splits No The paper is theoretical and does not involve empirical evaluation on datasets. Therefore, no information regarding dataset splits for training, validation, or testing is provided.
Hardware Specification No The paper is theoretical and does not describe running experiments. Therefore, no hardware specifications are mentioned.
Software Dependencies No The paper is theoretical and does not describe running experiments. While it discusses convex programming, it does not mention specific software or library dependencies with version numbers.
Experiment Setup No The paper is theoretical and does not describe running experiments. Therefore, no experimental setup details like hyperparameters or training configurations are provided.