A widely-used class of algorithms to understand data is hierarchical clustering, but it is often difficult to reconcile the results of these algorithms with hierarchies constructed by humans. Interaction, or querying humans for constraints on the data, is a popular solution for addressing this discrepancy. In this paper, we propose using leave-one-out interactions to achieve better hierarchies ...

I review the main steps made so far towards a detailed (semi) analytical model for the hierarchical clustering of bound virialized objects (i.e., haloes) in the gravitational instability scenario. I focus on those models relying on the spherical collapse approximation which have led to the most complete description. The work is divided in two parts: a first one dealing with the mass function of...

There are a large number of herbal gardens in India and most of them are not linked and classified the availability of medicinal plant species of these gardens. In this study, the hierarchical clustering technique of data mining was applied on the herbal gardens of India and classified in order to discover meaningful patterns such as what type of habit of medicinal plant species is present in w...

Given a p-dimensional proximity matrix Dp p, a sequence of correlation matrices, R = (R; R; : : :), is iteratively formed from it. Here R is the correlation matrix of the original proximity matrix D and R is the correlation matrix of R , n > 1. This sequence was rst introduced by McQuitty (1968), Breiger, Boorman and Arabie (1975) developed an algorithm, CONCOR, based on their rediscovery of it...

One of the most important issues in urban planning is developing sustainable public transportation. The basic condition for this purpose is analyzing current condition especially based on data. Data mining is a set of new techniques that are beyond statistical data analyzing. Clustering techniques is a subset of it that one of it’s techniques used for analyzing passengers’ trip. The result of...

Coordinated scan can gather host information for further attack more efficiently and stealthily than single-source port scans by distributing tasks amongst multiple sources. The existing coordinated scan detection methods are mostly based on scan behavior characteristics in temporal or spatial correlation. However, these detection methods can be easily evaded with the increasingly sophisticated...

A class-conditional hierarchical clustering framework has been used to generalize and improve previously proposed condensing schemes to obtain multiple prototype classifiers. The proposed method conveniently uses geometric properties and clusters to efficiently obtain reduced sets of prototypes that accurately represent the data while significantly keeping its discriminating power. The benefits...

DIVCLUS-T is a descendant hierarchical clustering algorithm based on a monothetic bipartitional approach allowing the dendrogram of the hierarchy to be read as a decision tree. We propose in this paper a new version of this method called C-DIVCLUS-T which is able to take contiguity constraints into account. We apply C-DIVCLUS-T to hydrological areas described by agricultural and environmental v...

The aim of this paper is to address the nature of hierarchical clustering problems in systems with very large numbers of entities, and to propose specific speed improvements in the clustering algorithm. The motivation for this theme arises from the challenge of visualising the geographic and logical distribution of many tens of thousands of distance-learning students at the UK's Open University...

We present an approach for the automatic induction of concept hierarchies from text collections. We propose a novel guided agglomerative hierarchical clustering algorithm exploiting a hypernym oracle to drive the clustering process. By inherently integrating the hypernym oracle into the clustering algorithm, we overcome two main problems of unsupervised clustering approaches relying on the dist...