Another class of nonlinear algorithms, exemplified by CART (Breiman, Friedman, Olshen, & Stone, 1984) and MARS (Friedman, 1990), generalizes classicaltechniques by partitioning the training data into non-overlapping regions and fitting separate models in each of the regions. These two classes of algorithms extendlinear techniques in essentially independent directions, thus it seems worthwhile to investigate algorithms that incorporate aspects of both approaches to model estimation. Such algorithms would be related to CART and MARS as multilayer neural networks are related to linear statistical techniques.

Automated monitoring of vigilance in attention intensive tasks such as air traffic control or sonar operation is highly desirable. As the operator monitorsthe instrument, the instrument would monitor the operator, insuring against lapses. We have taken a first step toward this goal by using feedforwardneural networks trained with backpropagation to interpret event related potentials (ERPs) and electroencephalogram (EEG) associated withperiods of high and low vigilance. The accuracy of our system on an ERP data set averaged over 28 minutes was 96%, better than the 83% accuracy obtained using linear discriminant analysis. Practical vigilance monitoring will require prediction over shorter time periods. We were able to average the ERP over as little as 2 minutes and still get 90% correct prediction of a vigilance measure. Additionally, we achieved similarly good performance using segments of EEG power spectrum as short as 56 sec.

Learning is posed as a problem of function estimation, for which two principles ofsolution are considered: empirical risk minimization and structural risk minimization. These two principles are applied to two different statements ofthe function estimation problem: global and local. Systematic improvements in prediction power are illustrated in application to zip-code recognition.

We derive criteria for training adaptive classifier networks to perform unsupervised dataanalysis. The first criterion turns a simple Gaussian classifier into a simple Gaussian mixture analyser. The second criterion, which is much more generally applicable, is based on mutual information.

G/SPLINES is an algorithm for building functional models of data. It uses genetic search to discover combinations of basis functions which are then used to build a least-squares regression model. Because it produces a population of models which evolve over time rather than a single model, it allows analysis not possible with other regression-based approaches. 1 INTRODUCTION G/SPLINES is a hybrid of Friedman's Multivariable Adaptive Regression Splines (MARS) algorithm (Friedman, 1990) with Holland's Genetic Algorithm (Holland, 1975). G/SPLINES has advantages over MARS in that it requires fewer least-squares computations, is easily extendable to non-spline basis functions, may discover models inaccessible to local-variable selection algorithms, and allows significantly larger problems to be considered. These issues are discussed in (Rogers, 1991). This paper begins with a discussion of linear regression models, followed by a description of the G/SPLINES algorithm, and finishes with a series of experiments illustrating its performance, robustness, and analysis capabilities.

Stephen M. Omohundro International Computer Science Institute 1947 CenteJ' Street, Suite 600 Berkeley, California 94704 Abstract "Best-first model merging" is a general technique for dynamically choosing the structure of a neural or related architecture while avoiding overfitting.It is applicable to both leaming and recognition tasks and often generalizes significantly better than fixed structures. We demonstrate theapproach applied to the tasks of choosing radial basis functions for function learning, choosing local affine models for curve and constraint surface modelling, and choosing the structure of a balltree or bumptree to maximize efficiency of access. 1 TOWARD MORE COGNITIVE LEARNING Standard backpropagation neural networks learn in a way which appears to be quite different fromhuman leaming. Viewed as a cognitive system, a standard network always maintains acomplete model of its domain. This model is mostly wrong initially, but gets gradually better and better as data appears. The net deals with all data in much the same way and has no representation for the strength of evidence behind a certain conclusion. The network architecture is usually chosen before any data is seen and the processing is much the same in the early phases of learning as in the late phases.

We compare two strategies for training connectionist (as well as nonconnectionist) modelsfor statistical pattern recognition. The probabilistic strategy is based on the notion that Bayesian discrimination (i.e.- optimal classification) isachieved when the classifier learns the a posteriori class distributions of the random feature vector. The differential strategy is based on the notion that the identity of the largest class a posteriori probability of the feature vector is all that is needed to achieve Bayesian discrimination. Each strategy is directly linked to a family ofobjective functions that can be used in the supervised training procedure. We prove that the probabilistic strategy - linked with error measure objective functions such as mean-squared-error and cross-entropy - typically used to train classifiers necessarily requires larger training sets and more complex classifier architectures than those needed to approximate the Bayesian discriminant function.In contrast.

The method of Structural Risk Minimization refers to tuning the capacity of the classifier to the available amount of training data.

Another class of nonlinear algorithms, exemplified by CART (Breiman, Friedman, Olshen, & Stone, 1984) and MARS (Friedman, 1990), generalizes classical techniques by partitioning the training data into non-overlapping regions and fitting separate models in each of the regions. These two classes of algorithms extend linear techniques in essentially independent directions, thus it seems worthwhile to investigate algorithms that incorporate aspects of both approaches to model estimation. Such algorithms would be related to CART and MARS as multilayer neural networks are related to linear statistical techniques. In this paper we present a candidate for such an algorithm. The algorithm that we present partitions its training data in the manner of CART or MARS, but it does so in a parallel, online manner that can be described as the stochastic optimization of an appropriate cost functional.

Three methods for improving the performance of (gaussian) radial basis function (RBF) networks were tested on the NETtaik task. In RBF, a new example is classified by computing its Euclidean distance to a set of centers chosen by unsupervised methods. The application of supervised learning to learn a non-Euclidean distance metric was found to reduce the error rate of RBF networks, while supervised learning of each center's variance resulted in inferior performance. The best improvement in accuracy was achieved by networks called generalized radial basis function (GRBF) networks. In GRBF, the center locations are determined by supervised learning. After training on 1000 words, RBF classifies 56.5% of letters correct, while GRBF scores 73.4% letters correct (on a separate test set). From these and other experiments, we conclude that supervised learning of center locations can be very important for radial basis function learning.