This paper presents a new learning algorithm based on decision tree ensembles. In opposition to the classical decision tree induction method, the trees of the ensemble are built by selecting the tests during their induction fully at random. This extreme

Diversity among the members of a team of classifiers is deemed to be a key issue in classifier combination. However, measuring diversity is not straightforward because there is no generally accepted formal definition. We have found and studied ten statistics which can measure diversity among binary classifier outputs (correct or incorrect vote for the class label): four averaged pairwise measures (the Q statistic, the correlation, the disagreement and the double fault) and six non-pairwise measures (the entropy of the votes, the difficulty index, the Kohavi-Wolpert variance, the interrater agreement, the generalized diversity, and the coincident failure diversity). Four experiments have been designed to examine the relationship between the accuracy of the team and the measures of diversity, and among the measures themselves. Although there are proven connections between diversity and accuracy in some special cases, our results raise some doubts about the usefulness of diversity measures in building classifier ensembles in real-life pattern recognition problems.

In this paper, we discuss round robin classification (aka pairwise classification), a technique for handling multi-class problems with binary classifiers by learning one classifier for each pair of classes. We present an empirical evaluation of the method, implemented as a wrapper around the Ripper rule learning algorithm, on 20 multi-class datasets from the UCI database repository. Our results show that the technique is very likely to improve Ripper's classification accuracy without having a high risk of decreasing it. More importantly, we give a general theoretical analysis of the complexity of the approach and show that its run-time complexity is below that of the commonly used one-against-all technique.

We present a method for constructing ensembles from libraries of thousands of models. Model libraries are generated using different learning algorithms and parameter settings. Forward stepwise selection is used to add to the ensemble the models that maximize its performance. Ensemble selection allows ensembles to be optimized to performance metric such as accuracy, cross entropy, mean precision, or ROC Area. Experiments with seven test problems and ten metrics demonstrate the benefit of ensemble selection.

Ensemble methods like bagging and boosting that combine the decisions of multiple hypotheses are some of the strongest existing machine learning methods. The diversity of the members of an ensemble is known to be an important factor in determining its generalization error. This paper presents a new method for generating ensembles that directly constructs diverse hypotheses using additional artificially-constructed training examples. The technique is a simple, general meta-learner that can use any strong learner as a base classifier to build diverse committees. Experimental results using decision-tree induction as a base learner demonstrate that this approach consistently achieves higher predictive accuracy than both the base classifier and bagging, and also obtains higher accuracy than boosting early in the learning curve when training data is limited.

Motivation: Genes and proteins are often associated with multiple names. More names are added as new functional or structural information is discovered. Because authors can use any one of the known names for a gene or protein, information retrieval and extraction would benefit from identifying the gene and protein terms that are synonyms of the same substance.

Abstract. Given an arbitrary data set, to which no particular parametrical, statistical or geometrical structure can be assumed, different clustering algorithms will in general produce different data partitions. In fact, several partitions can also be obtained by using a single clustering algorithm due to dependencies on initialization or the selection of the value of some design parameter. This paper addresses the problem of finding consistent clusters in data partitions, proposing the analysis of the most common associations performed in a majority voting scheme. Combination of clustering results are performed by transforming data partitions into a co-association sample matrix, which maps coherent associations. This matrix is then used to extract the underlying consistent clusters. The proposed methodology is evaluated in the context of k-means clustering, a new clustering algorithm- voting-k-means, being presented. Examples, using both simulated and real data, show how this majority voting combination scheme simultaneously handles the problems of selecting the number of clusters, and dependency on initialization. Furthermore, resulting clusters are not constrained to be hyper-spherically shaped. 1

The semantic Web Services vision requires that each service be annotated with semantic metadata. Manually creating such metadata is tedious and error-prone, and many software engineers, accustomed to tools that automatically generate WSDL, might not want to invest the additional e#ort. We therefore propose ASSAM, a tool that assists a user in creating semantic metadata for Web Services. ASSAM is intended for service consumers who want to integrate a number of services and therefore must annotate them according to some shared ontology. ASSAM is also relevant for service producers who have deployed a Web Service and want to make it compatible with an existing ontology. ASSAM's capabilities to automatically create semantic metadata are supported by two machine learning algorithms. First, we have developed an iterative relational classification algorithm for semantically classifying Web Services, their operations, and input and output messages. Second, to aggregate the data returned by multiple semantically related Web Services, we have developed a schema mapping algorithm that is based on an ensemble of string distance metrics.

Independence and dependence of classifier outputs have been debated in the recent literature giving rise to notions such as diversity, complementarity, orthogonality, etc. There seems to be no consensus on the meaning of these notions beyond the intuitive perception. Here we summarize 10 measures of classifier diversity: 4 pairwise and 6 non-pairwise measures. We derive the limits of the measures for 2 classifiers of equal accuracy.

Supporting continuous mining queries on data streams requires algorithms that (i) are fast, (ii) make light demands on memory resources, and (iii) are easily to adapt to concept drift. We propose a novel boosting ensemble method that achieves these objectives. The technique is based on a dynamic sample-weight assignment scheme that achieves the accuracy of traditional boosting without requiring multiple passes through the data. The technique assures faster learning and competitive accuracy using simpler base models. The scheme is then extended to handle concept drift via change detection. The change detection approach aims at significant data changes that could cause serious deterioration of the ensemble performance, and replaces the obsolete ensemble with one built from scratch. Experimental results confirm the advantages of our adaptive boosting scheme over previous approaches.