In an earlier paper, we introduced a new “boosting” algorithm called AdaBoost which, theoretically, can be used to significantly reduce the error of any learning algorithm that consistently generates classifiers whose performance is a little better than random guessing. We also introduced the related notion of a “pseudo-loss ” which is a method for forcing a learning algorithm of multi-label conceptsto concentrate on the labels that are hardest to discriminate. In this paper, we describe experiments we carried out to assess how well AdaBoost with and without pseudo-loss, performs on real learning problems. We performed two sets of experiments. The first set compared boosting to Breiman’s “bagging ” method when used to aggregate various classifiers (including decision trees and single attribute-value tests). We compared the performance of the two methods on a collection of machine-learning benchmarks. In the second set of experiments, we studied in more detail the performance of boosting using a nearest-neighbor classifier on an OCR problem.

The problem of combining preferences arises in several applications, such as combining the results of different search engines. This work describes an efficient algorithm for combining multiple preferences. We first give a formal framework for the problem. We then describe and analyze a new boosting algorithm for combining preferences called RankBoost. We also describe an efficient implementation of the algorithm for certain natural cases. We discuss two experiments we carried out to assess the performance of RankBoost. In the first experiment, we used the algorithm to combine different WWW search strategies, each of which is a query expansion for a given domain. For this task, we compare the performance of RankBoost to the individual search strategies. The second experiment is a collaborative-filtering task for making movie recommendations. Here, we present results comparing RankBoost to nearest-neighbor and regression algorithms.

Using an ensemble of classifiers, instead of a single classifier, can lead to improved generalization. The gains obtained by combining however, are often affected more by the selection of what is presented to the combiner, than by the actual combining method that is chosen. In this paper we focus on data selection and classifier training methods, in order to "prepare" classifiers for combining. We review a combining framework for classification problems that quantifies the need for reducing the correlation among individual classifiers. Then, we discuss several methods that make the classifiers in an ensemble more complementary. Experimental results are provided to illustrate the benefits and pitfalls of reducing the correlation among classifiers, especially when the training data is in limited supply. 2 1 Introduction A classifier's ability to meaningfully respond to novel patterns, or generalize, is perhaps its most important property (Levin et al., 1990; Wolpert, 1990). In...

Faces represent complex, multidimensional, meaningful visual stimuli and developing a computational model for face recognition is difficult [43]. We present a hybrid neural network solution which compares favorably with other methods. The system combines local image sampling, a self-organizing map neural network, and a convolutional neural network. The self-organizing map provides a quantization of the image samples into a topological space where inputs that are nearby in the original space are also nearby in the output space, thereby providing dimensionality reduction and invariance to minor changes in the image sample, and the convolutional neural network provides for partial invariance to translation, rotation, scale, and deformation. The convolutional network extracts successively larger features in a hierarchical set of layers. We present results using the Karhunen-Loeve transform in place of the self-organizing map, and a multi-layer perceptron in place of the convolutional netwo...

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.

Abstract not found

Multiple classifier fusion may generate more accurate classification than each of the constituent classifiers. Fusion is often based on fixed combination rules like the product and average. Only under strict probabilistic conditions can these rules be justified. We present here a simple rule for adapting the class combiner to the application. c decision templates (one per class) are estimated with the same training set that is used for the set of classifiers. These templates are then matched to the decision profile of new incoming objects by some similarity measure. We compare 11 versions of our model with 14 other techniques for classifier fusion on the Satimage and Phoneme datasets from the database ELENA. Our results show that decision templates based on integral type measures of similarity are superior to the other schemes on both data sets.

Based on negative correlation learning and evolutionary learning, this paper presents evolutionary ensembles with negative correlation learning (EENCL) to address the issues of automatic determination of the number of individual neural networks (NNs) in an ensemble and the exploitation of the interaction between individual NN design and combination. The idea of EENCL is to encourage different individual NNs in the ensemble to learn different parts or aspects of the training data so that the ensemble can better learn the entire training data. The cooperation and specialization among different individual NNs are considered during the individual NN design. This provides an opportunity for different NNs to interact with each other and to specialize. Experiments on two real-world problems demonstrate that EENCL can produce NN ensembles with good generalization ability. 1 Introduction Many real-world problems are too large and too complex for a single monolithic system to solve al...

This paper reviews research on combining artificial neural nets, and provides an overview of, and an introduction to, the papers contained this Special Issue, and its companion (Connection Science, 9, 1). Two main approaches, ensemble-based, and modular, are identified and considered. An ensemble, or committee, is made up of a set of nets, each of which is a general function approximator. The members of the ensemble are combined in order to obtain better generalisation performance than would be achieved by any of the individual nets. The main issues considered here under the heading of ensemble-based approaches, are (a) how to combine the outputs of the ensemble members (b) how to create candidate ensemble members and (c) which methods lead to the most effective ensembles? Under the heading of modular approaches we begin by considering a divide-and-conquer approach by which a function is automatically decomposed into a number of subfunctions which are treated by specialis...

Several researchers have experimentally shown that substantial improvements can be obtained in difficult pattern recognition problems by combining or integrating the outputs of multiple classifiers. This chapter provides an analytical framework to quantify the improvements in classification results due to combining. The results apply to both linear combiners and order statistics combiners. We first show that to a first order approximation, the error rate obtained over and above the Bayes error rate, is directly proportional to the variance of the actual decision boundaries around the Bayes optimum boundary. Combining classifiers in output space reduces this variance, and hence reduces the "added" error. If N unbiased classifiers are combined by simple averaging, the added error rate can be reduced by a factor of N if the individual errors in approximating the decision boundaries are uncorrelated. Expressions are then derived for linear combiners which are biased or correlated, and the effect of output correlations on ensemble performance is quantified. For order statistics based non-linear combiners, we derive expressions that indicate how much the median, the maximum and in general the ith order statistic can improve classifier performance. The analysis presented here facilitates the understanding of the relationships among error rates, classifier boundary distributions, and combining in output space. Experimental results on several public domain data sets are provided to illustrate the benefits of combining and to support the analytical results.