Filters








432,864 Hits in 4.0 sec

Finding Alternative Clusterings Using Constraints

Ian Davidson, Zijie Qi
2008 2008 Eighth IEEE International Conference on Data Mining  
In this work, we explore a general purpose approach to find an alternative clustering of the data with the aid of mustlink and cannot-link constraints.  ...  1 The aim of data mining is to find novel and actionable insights. However, most algorithms typically just find a single explanation of the data even though alternatives could exist.  ...  Conversely, you may also find the clustering found by A not particularly useful and actionable and wish to find an alternative to it.  ... 
doi:10.1109/icdm.2008.141 dblp:conf/icdm/DavidsonQ08 fatcat:fly4ek3q7rd3joqrpg4wlals4i

A principled and flexible framework for finding alternative clusterings

ZiJie Qi, Ian Davidson
2009 Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '09  
The problem of finding an alternative to a given original clustering has received little attention in the literature.  ...  In this work, we explore a principled and flexible framework in order to find alternative clusterings of the data.  ...  Acknowledgments The authors thank the anonymous reviewers for their excellent comments and the NSF for support of this work via GRANT IIS-0801528 CAREER: Knowledge Enhanced Clustering with Constraints.  ... 
doi:10.1145/1557019.1557099 dblp:conf/kdd/QiD09 fatcat:5yy47focjjbuxiqu32qnqr5nmm

Alternative Blockmodelling [article]

Oscar Correa and Jeffrey Chan and Vinh Nguyen
2019 arXiv   pre-print
Community finding field encompasses approaches which try to discover clusters where nodes are tightly related within them but loosely related with nodes of other clusters.  ...  Our methodology is presented through two approaches, (a) inclusion of cannot-link constraints and (b) dissimilarity between image matrices.  ...  In a second experiment they remove those constraints and find an alternative clustering.  ... 
arXiv:1908.02575v1 fatcat:mkz5dzjp2bfrxaxbsqik3rtdqy

How to "alternatize" a clustering algorithm

M. Shahriar Hossain, Naren Ramakrishnan, Ian Davidson, Layne T. Watson
2012 Data mining and knowledge discovery  
Given a clustering algorithm, how can we adapt it to find multiple, nonredundant, high-quality clusterings?  ...  Our framework works in both simultaneous and sequential learning formulations and can mine an arbitrary number of alternative clusterings.  ...  Acknowledgements This work is supported in part by the Institute for Critical Technology and Applied Science -Virginia Tech, the US National Science Foundation through grants CCF-0937133, AFRL through  ... 
doi:10.1007/s10618-012-0288-4 fatcat:ri6vjr2cqve2hg3vyxgepqieye

Data-driven solutions for building environmental impact assessment

Qifeng Zhou, Hao Zhou, Yimin Zhu, Tao Li
2015 Proceedings of the 2015 IEEE 9th International Conference on Semantic Computing (IEEE ICSC 2015)  
Second, a unified framework for solving constraint based clustering ensemble selection is proposed to extend the environmental impact assessment range from the building level to the regional level.  ...  Finally, a multiple disparate clustering method is presented to help sustainable new buildings design.  ...  clustering on the EI dataset; 4: Calculating the difference of two clustering results; 5: Finding the design alternatives.  ... 
doi:10.1109/icosc.2015.7050826 dblp:conf/semco/ZhouZZL15 fatcat:wapxjigp4jbyvhvlc72wmvdiba

Improving Alternative Text Clustering Quality in the Avoiding Bias Task with Spectral and Flat Partition Algorithms [chapter]

M. Eduardo Ares, Javier Parapar, Álvaro Barreiro
2010 Lecture Notes in Computer Science  
In this paper we put the focus on the quality of these alternative clusterings, proposing two approaches based in the use of negative constraints in conjunction with spectral clustering techniques.  ...  The problems of finding alternative clusterings and avoiding bias have gained popularity over the last years.  ...  Gondek and Hoffman introduced in [1] another strategy to find alternative clusters using Conditional Information Bottleneck clustering.  ... 
doi:10.1007/978-3-642-15251-1_32 fatcat:nxm7mnrcz5bl3inmvz443umdta

Clustering with Constraints [chapter]

Ian Davidson
2016 Encyclopedia of Database Systems  
Given a set of points P to cluster and a set of constraints C, the aim of clustering with constraints is to use the constraints to improve the clustering results.  ...  Semi-supervised clustering The area of clustering with constraints makes use of hints or advice in the form of constraints to aid or bias the clustering process.  ...  Gondek has explored using constraints to find orthogonal/alternative clusterings of data [11, 3] .  ... 
doi:10.1007/978-1-4899-7993-3_610-2 fatcat:y5tv4hgfavgpnmo46rzd2gqotq

COALA: A Novel Approach for the Extraction of an Alternate Clustering of High Quality and High Dissimilarity

Eric Bae, James Bailey
2006 IEEE International Conference on Data Mining. Proceedings  
However, traditional clustering methods concentrate on producing a single solution, even though multiple alternative clusterings may exist.  ...  This naturally leads to clustering solutions being highly dependent on the similarity function implemented by the particular algorithm used [16] .  ...  The technique uses the pre-defined class labels as additional information with which an alternate clustering is found.  ... 
doi:10.1109/icdm.2006.37 dblp:conf/icdm/BaeB06 fatcat:uk5awiin4nagbhir6o7q5cfmtm

Constrained Clustering and Multiple Kernel Learning without Pairwise Constraint Relaxation [article]

Benedikt Boecking, Vincent Jeanselme, Artur Dubrawski
2022 arXiv   pre-print
We introduce a new constrained clustering algorithm that jointly clusters data and learns a kernel in accordance with the available pairwise constraints.  ...  Clustering under pairwise constraints is an important knowledge discovery tool that enables the learning of appropriate kernels or distance metrics to improve clustering performance.  ...  Our proposed Sequential Model Based Optimization strategy finds good solutions quickly, but we also find that random parameter optimization over the sparse constrained space provides a good alternative  ... 
arXiv:2203.12546v1 fatcat:h43kwo6a6reyjj6kly4w7z5zu4

Scatter/Gather Clustering: Flexibly Incorporating User Feedback to Steer Clustering Results

M. Shahriar Hossain, Praveen Kumar Reddy Ojili, Cindy Grimm, Rolf Muller, Layne T. Watson, Naren Ramakrishnan
2012 IEEE Transactions on Visualization and Computer Graphics  
-451069, and DBI-1053171, the US Army Research Office (award id 451069), CRA Distributed Mentor Program, AFRL through grant FA8650-09-2-3938, and AFOSR through grant FA9550-09-1-0153.  ...  ACKNOWLEDGMENTS This work is supported in part by the Institute for Critical Technology and Applied Science -Virginia Tech, the US National Science Foundation through grants CCF-0937133, CCF-0702662, DBI  ...  In all the cases in addition to finding an alternative partitioning, the expert reported that the noise level was reduced in the alternative clustering.  ... 
doi:10.1109/tvcg.2012.258 pmid:26357192 fatcat:t5wuvtbbjvgudpww56fgzdfbru

SMVC

Stephan Günnemann, Ines Färber, Matthias Rüdiger, Thomas Seidl
2014 Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '14  
To better capture the data's complexity, methods aiming at the detection of multiple, alternative clusterings have been proposed.  ...  For efficient learning, we propose the algorithm SMVC using variational Bayesian methods.  ...  These constraints, however, are used for a different goal: they guide the clustering method to find a single new clustering.  ... 
doi:10.1145/2623330.2623734 dblp:conf/kdd/GunnemannFRS14 fatcat:ny3xg3m7urc3dcdbgi7dmjct3a

UNIFIED APPROACH TO DEPENDENT AND DISPARATE CLUSTERING OF NONHOMOGENOUS DATA

D.R. Easterling et al
2019 International Journal of Applied Mathematics  
We present an approach to cluster such nonhomogeneous datasets by using the relationships to impose either dependent clustering or disparate clustering constraints.  ...  This enables us to achieve dependent clustering and disparate clustering using the same optimization framework by merely maximizing versus minimizing the objective function.  ...  Finding alternative clusterings We investigate alternative clustering using the Portrait dataset as studied in [14] .  ... 
doi:10.12732/ijam.v32i3.3 fatcat:m25ohki2mrb7xa5tximy4lpxhq

Guided learning for role discovery (GLRD)

Sean Gilpin, Tina Eliassi-Rad, Ian Davidson
2013 Proceedings of the 19th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '13  
We provide an alternating least squares framework that allows convex constraints to be placed on the role discovery problem, which can provide useful supervision.  ...  In contrast to community discovery, which finds groups of highly connected nodes, the role discovery problem finds groups of nodes that share similar topological structure in the graph, and hence a common  ...  We also show how alternativeness can be used to find an alternative set of roles to the underlying community structure.  ... 
doi:10.1145/2487575.2487620 dblp:conf/kdd/GilpinED13 fatcat:nhrlgxeawfcrtoxdk3a5epmd6m

A Binary Optimization Approach for Constrained K-Means Clustering [article]

Huu Le, Anders Eriksson, Thanh-Toan Do, Michael Milford
2018 arXiv   pre-print
While the conventional Lloyd method, which alternates between centroid update and cluster assignment, is primarily used in practice, it may converge to a solution with empty clusters.  ...  This approach allows us to solve constrained K-Means where multiple types of constraints can be simultaneously enforced.  ...  Another type of constraint that finds its use in many clustering problems is the must-link (or cannot-link) constraints [23] .  ... 
arXiv:1810.10134v2 fatcat:suhnuvqncfaqtahshcd4pla3y4

Discovering Multiple Clustering Solutions: Grouping Objects in Different Views of the Data

Emmanuel Muller, Stephan Gunnemann, Ines Farber, Thomas Seidl
2010 2010 IEEE International Conference on Data Mining  
Most definitions provide only a single clustering solution For example, K -MEANS Aims at a single partitioning of the data Each object is assigned to exactly one cluster Aims at one clustering solution  ...  One set of K clusters forming the resulting groups of objects ⇒ In contrast, we focus on multiple clustering solutions...  ...  between clusterings, alternative should realize different density profile/histogram (Vinh & Epps, 2010) : based on conditional entropy, able to use a set of clusterings as input find (probabilistic)  ... 
doi:10.1109/icdm.2010.85 dblp:conf/icdm/MullerGFS10 fatcat:uaqjt4khojfcjdji34qwgpnn4y
« Previous Showing results 1 — 15 out of 432,864 results