We describe the RA scanner, a novel system for the examination of patients sufferingfrom rheumatoid arthritis. The RA scanner is based on a novel laser-based imaging technique which is sensitive to the optical characteristics of finger joint tissue. Based on the laser images, finger joints are classified according to whether the inflammatory status has improved or worsened. To perform the classification task, various linear andkernel-based systems were implemented and their performances were compared. Special emphasis was put on measures to reliably perform parametertuning and evaluation, since only a very small data set was available. Based on the results presented in this paper, it was concluded thatthe RA scanner permits a reliable classification of pathological finger joints, thus paving the way for a further development from prototype to product stage.

Most text categorization systems use simple models of documents and document collections. In this paper we describe a technique that improves asimple web page classifier's performance on pages from a new, unseen web site, by exploiting link structure within a site as well as page structure within hub pages. On real-world test cases, this technique significantly and substantially improves the accuracy of a bag-of-words classifier, reducing error rate by about half, on average. The system uses a variant of co-training to exploit unlabeled data from a new site. Pages are labeled using the base classifier; the results are used by a restricted wrapper-learner to propose potential "main-category anchor wrappers"; and finally, these wrappers are used as features by a third learner to find a categorization of the site that implies a simple hub structure, but which also largely agrees with the original bag-of-words classifier.

AdaBoost [3] minimizes an upper error bound which is an exponential function of the margin on the training set [14]. However, the ultimate goal in applications of pattern classification is always minimum error rate. On the other hand, AdaBoost needs an effective procedure for learning weak classifiers, which by itself is difficult especially for high dimensional data. In this paper, we present a novel procedure, called FloatBoost, for learning a better boosted classifier. FloatBoost uses a backtrack mechanism after each iteration of AdaBoost to remove weak classifiers which cause higher error rates. The resulting float-boosted classifier consists of fewer weak classifiers yet achieves lower error rates than AdaBoost in both training and test. We also propose a statistical model for learning weak classifiers, based on a stagewise approximation of the posterior using an overcomplete set of scalar features. Experimental comparisonsof FloatBoost and AdaBoost are provided through a difficult classification problem, face detection, where the goal is to learn from training examples a highly nonlinear classifier to differentiate between faceand nonface patterns in a high dimensional space. The results clearly demonstrate the promises made by FloatBoost over AdaBoost.

We introduce a general family of kernels based on weighted transducers orrational relations, rational kernels, that can be used for analysis of variable-length sequences or more generally weighted automata, in applications suchas computational biology or speech recognition. We show that rational kernels can be computed efficiently using a general algorithm ofcomposition of weighted transducers and a general single-source shortest-distance algorithm. We also describe several general families of positive definite symmetric rational kernels. These general kernels can be combined with Support Vector Machines to form efficient and powerful techniquesfor spoken-dialog classification: highly complex kernels become easy to design and implement and lead to substantial improvements inthe classification accuracy. We also show that the string kernels considered in applications to computational biology are all specific instances ofrational kernels.

For large, real-world inductive learning problems, the number of training examples often must be limited due to the costs associated with procuring, preparing, and storing the training examples and/or the computational costs associated with learning from them. In such circumstances, one question of practical importance is: if only n training examples can be selected, in what proportion should the classes be represented? In this article we help to answer this question by analyzing, for a fixed training-set size, the relationship between the class distribution of the training data and the performance of classification trees induced from these data. We study twenty-six data sets and, for each, determine the best class distribution for learning. The naturally occurring class distribution is shown to generally perform well when classifier performance is evaluated using undifferentiated error rate (0/1 loss). However, when the area under the ROC curve is used to evaluate classifier performance, a balanced distribution is shown to perform well. Since neither of these choices for class distribution always generates the best-performing classifier, we introduce a "budget-sensitive" progressive sampling algorithm for selecting training examples based on the class associated with each example. An empirical analysis of this algorithm shows that the class distribution of the resulting training set yields classifiers with good (nearly-optimal) classification performance.

The proliferation of online information sources has led to an increased use of wrappers for extracting data from Web sources. While most of the previous research has focused on quick and efficient generation of wrappers, the development of tools for wrapper maintenance has received less attention. This is an important research problem because Web sources often change in ways that prevent the wrappers from extracting data correctly. We present an efficient algorithm that learns structural information about data from positive examples alone. We describe how this information can be used for two wrapper maintenance applications: wrapper verification and reinduction. The wrapper verification system detects when a wrapper is not extracting correct data, usually because the Web source has changed its format. The reinduction algorithm automatically recovers from changes in the Web source by identifying data on Web pages so that a new wrapper may be generated for this source. To validate our approach, we monitored 27 wrappers over a period of a year. The verification algorithm correctly discovered 35 of the 37 wrapper changes, and made 16 mistakes, resulting in precision of 0.73 and recall of 0.95. We validated the reinduction algorithm on ten Web sources. We were able to successfully reinduce the wrappers, obtaining precision and recall values of 0.90 and 0.80 on the data extraction task.

This paper presents a method for obtaining class membership probability estimates for multiclass classification problems by coupling the probability estimates produced by binary classifiers. This is an extension for arbitrary code matrices of a method due to Hastie and Tibshirani for pairwise coupling of probability estimates. Experimental results with Boosted Naive Bayes show that our method produces calibrated class membership probability estimates, while having similar classification accuracy as loss-based decoding, a method for obtaining the most likely class that does not generate probability estimates.

This paper develops a new approach for extremely fast detection in domains where the distribution of positive and negative examples is highly skewed (e.g.

When designing a two-alternative classifier, one ordinarily aims to maximize the classifier's ability to discriminate between members of the two classes. We describe a situation in a real-world business application of machine-learning prediction in which an additional constraint is placed on the nature of the solution: that the classifier achieve a specified correct acceptance or correct rejection rate (i.e., that it achieve a fixed accuracy on members of one class or the other). Our domain is predicting churn in the telecommunications industry. Churn refers to customers who switch from one service provider to another. We propose four algorithms for training a classifier subject to this domain constraint, and present results showing that each algorithm yields a reliable improvement in performance.

In this paper, we study bootstrapping algorithms for learning from unlabeled data. The general idea in bootstrapping is to use some initial labeled data to build a (possibly partial) predictive labeling procedure; then use the labeling procedure to label more data; then use the newly labeled data to build a new predictive procedure and so on. This process can be iterated until a fixed point is reached or some other stopping criterion is met. Here we give P AC style bounds on generalization error which can be used to formally justify certain boostrapping algorithms. One well-known form of bootstrapping is the EM algorithm (Dempster, Laird and Rubin, 1977).