We develop a common theoretical framework for combining classifiers which use distinct pattern representations and show that many existing schemes can be considered as special cases of compound classification where all the pattern representations are used jointly to make a decision. An experimental comparison of various classifier combination schemes demonstrates that the combination rule developed under the most restrictive assumptions—the sum rule—outperforms other classifier combinations schemes. A sensitivity analysis of the various schemes to estimation errors is carried out to show that this finding can be justified theoretically.

Much of previous attention on decision trees focuses on the splitting criteria and optimization of tree sizes. The dilemma between overfitting and achieving maximum accuracy is seldom resolved. We propose a method to construct a decision tree based classifier that maintains highest accuracy on training data and improves on generalization accuracy as it grows in complexity. The classifier consists of multiple trees constructed systematically by pseudorandomly selecting subsets of components of the feature vector, that is, trees constructed in randomly chosen subspaces. The subspace method is compared to single-tree classifiers and other forest construction methods by experiments on publicly available datasets, where the method's superiority is demonstrated. We also discuss independence between trees in a forest and relate that to the combined classification accuracy. keywords: pattern recognition, decision tree, decision forest, stochastic discrimination, decision combination, classif...

Abstract—A wide variety of systems requires reliable personal recognition schemes to either confirm or determine the identity of an individual requesting their services. The purpose of such

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...

User verification systems that use a single biometric indicator often have to contend with noisy sensor data, restricted degrees of freedom, non-universality of the biometric trait and unacceptable error rates. Attempting to improve the performance of individual matchers in such situations may not prove to be effective because of these inherent problems. Multibiometric systems seek to alleviate some of these drawbacks by providing multiple evidences of the same identity. These systems help achieve an increase in performance that may not be possible using a single biometric indicator. Further, multibiometric systems provide anti-spoofing measures by making it difficult for an intruder to spoof multiple biometric traits simultaneously. However, an effective fusion scheme is necessary to combine the information presented by multiple domain experts. This paper addresses the problem of information fusion in biometric verification systems by combining information at the matching score level. Experimental results on combining three biometric modalities (face, fingerprint and hand geometry) are presented.

Stacked generalization is a general method of using a high-level model to combine lower-level models to achieve greater predictive accuracy. In this paper we address two crucial issues which have been considered to be a `black art' in classification tasks ever since the introduction of stacked generalization in 1992 by Wolpert: the type of generalizer that is suitable to derive the higher-level model, and the kind of attributes that should be used as its input. We find that best results are obtained when the higher-level model combines the confidence (and not just the predictions) of the lower-level ones.

Whereas most state-of-the-art speech recognition systems use spectral or cepstral representations of the speech signal, there have also been some promising attempts at using articulatory information. These attempts have been motivated by two major assumptions: first, coarticulation can be modeled more naturally due to the inherently asynchronous nature of articulatory information. Second, it is assumed that the overall patterns in the speech signal caused by articulatory gestures are more robust to noise and speaker-dependent acoustic variation than spectral parameters. A third assumption can be made, viz. that acoustic and articulatory representations of speech can supply mutually complementary information to a speech recognizer, in which case the combination of these representations might be beneficial. Previously, articulatory-based speech recognizers have exclusively been developed for clean speech; the potential of an articulatory representation of the speech signal for noisy test...

We introduce a very large family of binary features for two-dimensional shapes. The salient ones for separating particular shapes are determined by inductive learning during the construction of classi cation trees. There is a feature for every possible geometric arrangement of local topographic codes. The arrangements express coarse constraints on relative angles and distances among the code locations and are nearly invariant to substantial a ne and non-linear deformations. They are also partially ordered, which makes it possible to narrow the search for informative ones at each node of the tree. Di erent trees correspond to di erent aspects of shape. They are statistically weakly dependent due to randomization and are aggregated in a simple way. Adapting the algorithm to a shape family is then fully automatic once training samples are provided. As an illustration, we classify handwritten digits from the NIST database � the error rate is:7%.

The latest research indicates using a combination of biometric avenues for human identification is more effective, and far more challenging.

When more than a single classifier has been trained for the same recognition problem the question arises how this set of classifiers may be combined into a final decision rule. Several fixed combining rules are used that depend on the output values of the base classifiers only. They are almost always suboptimal.