A new approach to isotonic agglomerative hierarchical clustering

Hierarchical clustering methods must be isotonic for the construction of ultrametric. We present a general strategy to widen the class of isotonic methods implemented by agglomerative algorithms. At each step of the agglomeration we allow one of several admissible pairs to be chosen. Then under mild assumptions an appropriate definition of admissibility guarantees isotony. Moreover we consider the use of the new methods to compute locally optimal ultrametrics. Two examples demonstrate the ability to define new agglomerative methods superior to their traditional competitors.     

Lowdimensional Additive Overlapping Clustering

To reveal the structure underlying two-way two-mode object by variable data, Mirkin (1987) has proposed an additive overlapping clustering model. This model implies an overlapping clustering of the objects and a reconstruction of the data, with the reconstructed variable profile of an object being a summation of the variable profiles of the clusters it belongs to. Grasping the additive (overlapping) clustering structure of object by variable data may, however, be seriously hampered in case the data include a very large number of variables. To deal with this problem, we propose a new model that simultaneously clusters the objects in overlapping clusters and reduces the variable space; as such, the model implies that the cluster profiles and, hence, the reconstructed data profiles are constrained to lie in a lowdimensional space. An alternating least squares (ALS) algorithm to fit the new model to a given data set will be presented, along with a simulation study and an illustrative example that makes use of empirical data.     

A Copula-Based Algorithm for Discovering Patterns of Dependent Observations

The main aim of this work is the study of clustering dependent data by means of copula functions. Copulas are popular multivariate tools whose importance within clustering methods has not been investigated yet in detail. We propose a new algorithm (CoClust in brief) that allows to cluster dependent data according to the multivariate structure of the generating process without any assumption on the margins. Moreover, the approach does not require either to choose a starting classification or to set a priori the number of clusters; in fact, the CoClust selects them by using a criterion based on the log–likelihood of a copula fit. We test our proposal on simulated data for different dependence scenarios and compare it with a model–based clustering technique. Finally, we show applications of the CoClust to real microarray data of breast-cancer patients.     

A Criterion Based on the Mahalanobis Distance for Cluster Analysis with Subsampling

A two-level data set consists of entities of a higher level (say populations), each one being composed of several units of the lower level (say individuals). Observations are made at the individual level, whereas population characteristics are aggregated from individual data. Cluster analysis with subsampling of populations is a cluster analysis based on individual data that aims at clustering populations rather than individuals. In this article, we extend existing optimality criteria for cluster analysis with subsampling of populations to deal with situations where population characteristics are not the mean of individual data. A new criterion that depends on the Mahalanobis distance is also defined. The criteria are compared using simulated examples and an ecological data set of tree species in a tropical rain forest.     

A Survey on Feature Weighting Based K-Means Algorithms

In a real-world data set, there is always the possibility, rather high in our opinion, that different features may have different degrees of relevance. Most machine learning algorithms deal with this fact by either selecting or deselecting features in the data preprocessing phase. However, we maintain that even among relevant features there may be different degrees of relevance, and this should be taken into account during the clustering process.With over 50 years of history, K-Means is arguably the most popular partitional clustering algorithm there is. The first K-Means based clustering algorithm to compute feature weights was designed just over 30 years ago. Various such algorithms have been designed since but there has not been, to our knowledge, a survey integrating empirical evidence of cluster recovery ability, common flaws, and possible directions for future research. This paper elaborates on the concept of feature weighting and addresses these issues by critically analyzing some of the most popular, or innovative, feature weighting mechanisms based in K-Means.     

Classification of Asymmetric Proximity Data

When clustering asymmetric proximity data, only the average amounts are often considered by assuming that the asymmetry is due to noise. But when the asymmetry is structural, as typically may happen for exchange flows, migration data or confusion data, this may strongly affect the search for the groups because the directions of the exchanges are ignored and not integrated in the clustering process. The clustering model proposed here relies on the decomposition of the asymmetric dissimilarity matrix into symmetric and skew-symmetric effects both decomposed in within and between cluster effects. The classification structures used here are generally based on two different partitions of the objects fitted to the symmetric and the skew-symmetric part of the data, respectively; the restricted case is also presented where the partition fits jointly both of them allowing for clusters of objects similar with respect to the average amounts and directions of the data. Parsimonious models are presented which allow for effective and simple graphical representations of the results.     

A permutation-based algorithm for block clustering

Hartigan (1972) discusses the direct clustering of a matrix of data into homogeneous blocks. He introduces a stepwise divisive method for block clustering within a certain class of block structures which induce clustering trees for both row and column margins. While this class of structures is appealing, the stopping criterion for his method, which is based on asymptotic theory and the assumption that the individual elements of the data matrix are normally distributed, is quite restrictive. In this paper we propose a permutation-based algorithm for block clustering within the same class of block structures. By using permutation arguments to decide where to split and when to stop, our algorithm becomes applicable in a wide variety of cases, including matrices of categorical data and matrices of small-to-moderate size. In addition, our algorithm offers considerable flexibility in how block homogeneity is defined. The algorithm is studied in a series of simulation experiments on matrices of known structure, and illustrated in examples drawn from the fields of taxonomy, political science, and data architecture.     

Recent convergence results for the fuzzy c-means clustering algorithms

One of the main techniques embodied in many pattern recognition systems is cluster analysis — the identification of substructure in unlabeled data sets. The fuzzy c-means algorithms (FCM) have often been used to solve certain types of clustering problems. During the last two years several new local results concerning both numerical and stochastic convergence of FCM have been found. Numerical results describe how the algorithms behave when evaluated as optimization algorithms for finding minima of the corresponding family of fuzzy c-means functionals. Stochastic properties refer to the accuracy of minima of FCM functionals as approximations to parameters of statistical populations which are sometimes assumed to be associated with the data. The purpose of this paper is to collect the main global and local, numerical and stochastic, convergence results for FCM in a brief and unified way.     

Space-conserving agglomerative algorithms

This paper evaluates a general, infinite family of clustering algorithms, called the Lance and Williams algorithms, with respect to the space-conserving criterion. An admissible clustering criterion is defined using the space conserving idea. Necessary and sufficient conditions for Lance and Williams clustering algorithms to satisfy space-conserving admissibility are provided. Space-dilating, space-contracting, and well-structured clustering algorithms are also discussed.The work of J. Van Ness was supported by NSF Grant #DMS 9201075.     

A Note on K-modes Clustering

Recently, Chaturvedi, Green and Carroll (2001) presented a nonparametric approach to deriving clusters from categorical data using a new clustering procedure called K-modes. Huang (1998) proposed the K-modes clustering algorithm. In this note, we demonstrate the equivalence of the two K-modes procedures.     

Direct multicriteria clustering algorithms

In a multicriteria clustering problem, optimization over more than one criterion is required. The problem can be treated in different ways: by reduction to a clustering problem with the single criterion obtained as a combination of the given criteria; by constrained clustering algorithms where a selected critetion is considered as the clustering criterion and all others determine the constraints; or by direct algorithms. In this paper two types of direct algorithms for solving multicriteria clustering problem are proposed: the modified relocation algorithm, and the modified agglomerative algorithm. Different elaborations of these two types of algorithms are discussed and compared. Finally, two applications of the proposed algorithms are presented. Elaborated version of the talks presented at the First Conference of the International Federation of Classification Societies, Aachen, 1987, at the International Conference on Social Science Methodology, Dubrovnik, 1988, and at the Second Conference of the International Federation of Classification Societies, Charlottesville, 1989. This work was supported in part by the Research Council of Slovenia.     

Design of Blurring Mean-Shift Algorithms for Data Classification

The mean-shift algorithm is an iterative method of mode seeking and data clustering based on the kernel density estimator. The blurring mean-shift is an accelerated version which uses the original data only in the first step, then re-smoothes previous estimates. It converges to local centroids, but may suffer from problems of asymptotic bias, which fundamentally depend on the design of its smoothing components. This paper develops nearest-neighbor implementations and data-driven techniques of bandwidth selection, which enhance the clustering performance of the blurring method. These solutions can be applied to the whole class of mean-shift algorithms, including the iterative local mean method. Extended simulation experiments and applications to well known data-sets show the goodness of the blurring estimator with respect to other algorithms.     

Robust Double Clustering: A Method Based on Alternating Concentration Steps

We propose two algorithms for robust two-mode partitioning of a data matrix in the presence of outliers. First we extend the robust k-means procedure to the case of biclustering, then we slightly relax the definition of outlier and propose a more flexible and parsimonious strategy, which anyway is inherently less robust. We discuss the breakdown properties of the algorithms, and illustrate the methods with simulations and three real examples. The author is grateful to four referees for detailed suggestions that led to an improved paper, and to Professor Vichi for support and careful reading of a first draft. Acknowledgements go also to Francesca Martella for advice.     

Variable Selection for Clustering and Classification

As data sets continue to grow in size and complexity, effective and efficient techniques are needed to target important features in the variable space. Many of the variable selection techniques that are commonly used alongside clustering algorithms are based upon determining the best variable subspace according to model fitting in a stepwise manner. These techniques are often computationally intensive and can require extended periods of time to run; in fact, some are prohibitively computationally expensive for high-dimensional data. In this paper, a novel variable selection technique is introduced for use in clustering and classification analyses that is both intuitive and computationally efficient. We focus largely on applications in mixture model-based learning, but the technique could be adapted for use with various other clustering/classification methods. Our approach is illustrated on both simulated and real data, highlighted by contrasting its performance with that of other comparable variable selection techniques on the real data sets.     

Constrained clustering and Kohonen Self-Organizing Maps

The Self-Organizing Feature Maps (SOFM; Kohonen 1984) algorithm is a well-known example of unsupervised learning in connectionism and is a clustering method closely related to the k-means. Generally the data set is available before running the algorithm and the clustering problem can be approached by an inertia criterion optimization. In this paper we consider the probabilistic approach to this problem. We propose a new algorithm based on the Expectation Maximization principle (EM; Dempster, Laird, and Rubin 1977). The new method can be viewed as a Kohonen type of EM and gives a better insight into the SOFM according to constrained clustering. We perform numerical experiments and compare our results with the standard Kohonen approach.     

MCS: A Method for Finding the Number of Clusters

This paper proposes a maximum clustering similarity (MCS) method for determining the number of clusters in a data set by studying the behavior of similarity indices comparing two (of several) clustering methods. The similarity between the two clusterings is calculated at the same number of clusters, using the indices of Rand (R), Fowlkes and Mallows (FM), and Kulczynski (K) each corrected for chance agreement. The number of clusters at which the index attains its maximum is a candidate for the optimal number of clusters. The proposed method is applied to simulated bivariate normal data, and further extended for use in circular data. Its performance is compared to the criteria discussed in Tibshirani, Walther, and Hastie (2001). The proposed method is not based on any distributional or data assumption which makes it widely applicable to any type of data that can be clustered using at least two clustering algorithms.     

Optimal Variable Weighting for Ultrametric and Additive Trees and K-means Partitioning: Methods and Software

K -means partitioning. We also describe some new features and improvements to the algorithm proposed by De Soete. Monte Carlo simulations have been conducted using different error conditions. In all cases (i.e., ultrametric or additive trees, or K-means partitioning), the simulation results indicate that the optimal weighting procedure should be used for analyzing data containing noisy variables that do not contribute relevant information to the classification structure. However, if the data involve error-perturbed variables that are relevant to the classification or outliers, it seems better to cluster or partition the entities by using variables with equal weights. A new computer program, OVW, which is available to researchers as freeware, implements improved algorithms for optimal variable weighting for ultrametric and additive tree clustering, and includes a new algorithm for optimal variable weighting for K-means partitioning.     

Variable selection in clustering

Standard clustering algorithms can completely fail to identify clear cluster structure if that structure is confined to a subset of the variables. A forward selection procedure for identifying the subset is proposed and studied in the context of complete linkage hierarchical clustering. The basic approach can be applied to other clustering methods, too.     

Fast random generation of binary,t-ary and other types of trees

Trees, and particularly binary trees, appear frequently in the classification literature. When studying the properties of the procedures that fit trees to sets of data, direct analysis can be too difficult, and Monte Carlo simulations may be necessary, requiring the implementation of algorithms for the generation of certain families of trees at random. In the present paper we use the properties of Prufer's enumeration of the set of completely labeled trees to obtain algorithms for the generation of completely labeled, as well as terminally labeled t-ary (and in particular binary) trees at random, i.e., with uniform distribution. Actually, these algorithms are general in that they can be used to generate random trees from any family that can be characterized in terms of the node degrees. The algorithms presented here are as fast as (in the case of terminally labeled trees) or faster than (in the case of completely labeled trees) any other existing procedure, and the memory requirements are minimal. Another advantage over existing algorithms is that there is no need to store pre-calculated tables.