Multi-way clustering and biclustering by the Ratio cut and Normalized cut in graphs

  title={Multi-way clustering and biclustering by the Ratio cut and Normalized cut in graphs},
  author={Neng Fan and Panos M. Pardalos},
  journal={Journal of Combinatorial Optimization},
  • Neng Fan, P. Pardalos
  • Published 1 February 2012
  • Computer Science
  • Journal of Combinatorial Optimization
In this paper, we consider the multi-way clustering problem based on graph partitioning models by the Ratio cut and Normalized cut. We formulate the problem using new quadratic models. Spectral relaxations, new semidefinite programming relaxations and linearization techniques are used to solve these problems. It has been shown that our proposed methods can obtain improved solutions. We also adapt our proposed techniques to the bipartite graph partitioning problem for biclustering. 

A Complementary Column Generation Approach for the Graph Equipartition Problem

This paper presents a mathematical programming model and proposes a complementary column generation approach to solve the problem of partitioning a complete weighted graph into complete subgraphs, each having the same number of vertices, with the objective of minimizing the sum of edge weights of the resulting sub graphs.

Improved linearized models for graph partitioning problem under capacity constraints

We investigate a variant of the Graph Partitioning Problem with capacity constraints imposed on the clusters, giving rise to quadratic constraints in 0–1 formulations. Several compact linearized

On the Two-Stage Stochastic Graph Partitioning Problem

This paper presents the stochastic mixed integer programming formulation for this problem with finite explicit scenarios and an equivalent integer linear programming formulation where some binary variables are relaxed to continuous ones.

Guaranteed clustering and biclustering via semidefinite programming

It is shown that the correct partition of the objects and features can be recovered from the optimal solution of a semidefinite program in the case that the given data consists of several disjoint sets of objects exhibiting similar features.

Combinatorial and nonlinear optimization methods with applications in data clustering, biclustering and power systems

This dissertation presents Combinatorial and Nonlinear Optimization Methods with Applications in Data Clustering, Biclustering and Power Systems, which combines combinatorics and graph theory.

Graph clustering with Boltzmann machines

Two mathematical programming formulations, the distance minimization and K-medoids, yield clusters of superior quality to those obtained with the Louvain algorithm, and establish the superiority of the Boltzmann machine over the traditional exact solver.

Normalized cuts clustering with prior knowledge and a pre-clustering stage

This work proposes an alternative to solve the normalized cuts problem for clustering, achieving same results as conventional spectral meth- ods but spending less processing time, and reaches comparable results with a lower computational cost.

An Improvement of Spectral Clustering via Message Passing and Density Sensitive Similarity

An algorithm on the basis of message passing, which makes use of a density adaptive similarity measure, describing the relations between data points and obtaining high-quality cluster centers through message passing mechanism in AP clustering is presented.



Integer Programming of Biclustering Based on Graph Models

It is proved that the spectral biclustering for Ratio cut and Normalized cut are the relaxation forms of these integer programming models, and also the Minmax cut for bic Lustering is equivalent to Normalized Cut for bIClustering.

Spectral K-Way Ratio-Cut Partitioning and Clustering

A spectral approach to multiway ratio-cut partitioning is developed which provides a generalization of the ratio- cut cost metric to k-way partitioning and a lower bound on this cost metric.

Graph partitioning using linear and semidefinite programming

Both linear and semidefinite relaxations for the equipartition problem are considered and numerical results on real data from France Telecom networks with up 900 nodes, and also on randomly generated problems are presented.

Linear and quadratic programming approaches for the general graph partitioning problem

The graph partitioning problem is formulated as a zero-one quadratic programming problem without the input of cardinalities and several new methods to determine the number of subsets and the cardinalities are presented for practical applications.

A min-max cut algorithm for graph partitioning and data clustering

This paper proposes a new algorithm for graph partitioning with an objective function that follows the min-max clustering principle, and demonstrates that a linearized search order based on linkage differential is better than that based on the Fiedler vector, providing another effective partitioning method.

Improved approximation algorithms for MAXk-CUT and MAX BISECTION

Polynomial-time approximation algorithms with nontrivial performance guarantees are presented for the problems of (a) partitioning the vertices of a weighted graph intok blocks so as to maximize the

Bipartite graph partitioning and data clustering

A new data clustering method based on partitioning the underlying bipartite graph, which shows that an approximate solution to the minimization problem can be obtained by computing a partial singular value decomposition of the associated edge weight matrix of the bipartites graph.

Multiset graph partitioning

Local optimality conditions are given for a quadratic programming formulation of the multiset graph partitioning problem related to the structure of the graph and properties of the weights.

New spectral methods for ratio cut partitioning and clustering

  • L. HagenA. Kahng
  • Computer Science
    IEEE Trans. Comput. Aided Des. Integr. Circuits Syst.
  • 1992
It is shown that the second smallest eigenvalue of a matrix derived from the netlist gives a provably good approximation of the optimal ratio cut partition cost.

Bipartite isoperimetric graph partitioning for data co-clustering

This paper proposes Isoperimetric Co-clustering Algorithm (ICA)—a new method for partitioning the bipartite graph that requires a simple solution to a sparse system of linear equations instead of the eigenvalue or SVD problem in the popular spectral co-cluster approach.