In the feature subset selection problem, a learning algorithm is faced with the problem of selecting a relevant subset of features upon which to focus its attention, while ignoring the rest. To achieve the best possible performance with a particular learning algorithm on a particular training set, a feature subset selection method should consider how the algorithm and the training set interact. We explore the relation between optimal feature subset selection and relevance. Our wrapper method searches for an optimal feature subset tailored to a particular algorithm and a domain. We study the strengths and weaknesses of the wrapper approach and show a series of improved designs. We compare the wrapper approach to induction without feature subset selection and to Relief, a filter approach to feature subset selection. Significant improvement in accuracy is achieved for some datasets for the two families of induction algorithms used: decision trees and

Methods for voting classification algorithms, such as Bagging and AdaBoost, have been shown to be very successful in improving the accuracy of certain classifiers for artificial and real-world datasets. We review these algorithms and describe a large empirical study comparing several variants in conjunction with a decision tree inducer (three variants) and a Naive-Bayes inducer. The purpose of the study is to improve our understanding of why and when these algorithms, which use perturbation, reweighting, and combination techniques, affect classification error. We provide a bias and variance decomposition of the error to show how different methods and variants influence these two terms. This allowed us to determine that Bagging reduced variance of unstable methods, while boosting methods (AdaBoost and Arc-x4) reduced both the bias and variance of unstable methods but increased the variance for Naive-Bayes, which was very stable. We observed that Arc-x4 behaves differently than AdaBoost if reweighting is used instead of resampling, indicating a fundamental difference. Voting variants, some of which are introduced in this paper, include: pruning versus no pruning, use of probabilistic estimates, weight perturbations (Wagging), and backfitting of data. We found that Bagging improves when probabilistic estimates in conjunction with no-pruning are used, as well as when the data was backfit. We measure tree sizes and show an interesting positive correlation between the increase in the average tree size in AdaBoost trials and its success in reducing the error. We compare the mean-squared error of voting methods to non-voting methods and show that the voting methods lead to large and significant reductions in the mean-squared errors. Practical problems that arise in implementing boosting algorithms are explored, including numerical instabilities and underflows. We use scatterplots that graphically show how AdaBoost reweights instances, emphasizing not only "hard" areas but also outliers and noise.

Abstract. Bagging and boosting are methods that generate a diverse ensemble of classifiers by manipulating the training data given to a “base ” learning algorithm. Breiman has pointed out that they rely for their effectiveness on the instability of the base learning algorithm. An alternative approach to generating an ensemble is to randomize the internal decisions made by the base algorithm. This general approach has been studied previously by Ali and Pazzani and by Dietterich and Kong. This paper compares the effectiveness of randomization, bagging, and boosting for improving the performance of the decision-tree algorithm C4.5. The experiments show that in situations with little or no classification noise, randomization is competitive with (and perhaps slightly superior to) bagging but not as accurate as boosting. In situations with substantial classification noise, bagging is much better than boosting, and sometimes better than randomization.

We analyze critically the use of classification accuracy to compare classifiers on natural data sets, providing a thorough investigation using ROC analysis, standard machine learning algorithms, and standard benchmark data sets. The results raise serious concerns about the use of accuracy for comparing classifiers and drawinto question the conclusions that can be drawn from such studies. In the course of the presentation, we describe and demonstrate what we believe to be the proper use of ROC analysis for comparative studies in machine learning research. We argue that this methodology is preferable both for making practical choices and for drawing scientific conclusions.

In real-world environments it is usually difficult to specify target operating conditions precisely. This uncertainty makes building robust classification systems problematic. We present a method for the comparison of classifier performance that is robust to imprecise class distributions and misclassification costs. The ROC convex hull method combines techniques from ROC analysis, decision analysis and computational geometry, and adapts them to the particulars of analyzing learned classifiers. The method is efficient and incremental, minimizes the management of classifier performance data, and allows for clear visual comparisons and sensitivity analyses. We then show that it is possible to build a hybrid classifier that will perform at least as well as the best available classifier for any target conditions. This robust performance extends across a wide variety of comparison frameworks, including the optimization of metrics such as accuracy, expected cost, lift, precision, recall, and ...

Deduplication is a key operation in integrating data from multiple sources. The main challenge in this task is designing a function that can resolve when a pair of records refer to the same entity in spite of various data inconsistencies. Most existing systems use hand-coded functions. One way to overcome the tedium of hand-coding is to train a classifier to distinguish between duplicates and non-duplicates. The success of this method critically hinges on being able to provide a covering and challenging set of training pairs that bring out the subtlety of the deduplication function. This is non-trivial because it requires manually searching for various data inconsistencies between any two records spread apart in large lists. We present our design of a learning-based deduplication system that uses a novel method of interactively discovering challenging training pairs using active learning. Our experiments on real-life datasets show that active learning signicantly reduces the number of instances needed to achieve high accuracy. We investigate various design issues that arise in building a system to provide interactive response, fast convergence, and interpretable output.

The "minimum margin" of an ensemble classifier on a given training set is, roughly speaking, the smallest vote it gives to any correct training label. Recent work has shown that the Adaboost algorithm is particularly effective at producing ensembles with large minimum margins, and theory suggests that this may account for its success at reducing generalization error. We note, however, that the problem of finding good margins is closely related to linear programming, and we use this connection to derive and test new "LPboosting" algorithms that achieve better minimum margins than Adaboost. However, these algorithms do not always yield better generalization performance. In fact, more often the opposite is true. We report on a series of controlled experiments which show that no simple version of the minimum-margin story can be complete. We conclude that the crucial question as to why boosting works so well in practice, and how to further improve upon it, remains mostly open. Some of our ...

Abstract not found

. We describe an experimental study of pruning methods for decision tree classifiers when the goal is minimizing loss rather than error. In addition to two common methods for error minimization, CART's cost-complexity pruning and C4.5's error-based pruning, we study the extension of cost-complexity pruning to loss and one pruning variant based on the Laplace correction. We perform an empirical comparison of these methods and evaluate them with respect to loss. We found that applying the Laplace correction to estimate the probability distributions at the leaves was beneficial to all pruning methods. Unlike in error minimization, and somewhat surprisingly, performing no pruning led to results that were on par with other methods in terms of the evaluation criteria. The main advantage of pruning was in the reduction of the decision tree size, sometimes by a factor of ten. While no method dominated others on all datasets, even for the same domain different pruning mechanisms are better for ...

MineSet TM , Silicon Graphics' interactive system for data mining, integrates three powerful technologies: database access, analytical data mining, and data visualization. It supports the knowledge discovery process from data access and preparation through iterative analysis and visualization to deployment. MineSet is based on a client-server architecture that scales to large databases. The database access component provides a rich set of operators that can be used to preprocess and transform the stored data into forms appropriate for visualization and analytical mining. The 3D visualization capabilities allow direct data visualization for exploratory analysis, including tools for displaying high-dimensional data containing geographical and hierarchical information. The analytical mining algorithms help identify potentially interesting models of the data, which can be viewed using visualization tools specialized for the learned models. Third party vendors can interface to the MineSet...