Human error is an Criticism should not be querulous, and umes of fast-changing increasingly important wasting, all knife and root puller, but guiding, sensory data that and addressable instructive, inspiring, a South wind, one needs to process concern in modernday not an East wind. Most institutions), and the automation that technology represents accidents waiting to surrounds us (for example, unfriendly computers happen. For example, in the Challenger explosion, We get by because humans excel at coping. the shortcomings of the O-rings had been High-technology accidents occur because known for several years. What feedback hundreds of alarms simultaneously all contributed strategy (for example, story telling, first-principle to the disaster. Likewise, when the lecturing) will most constructively correct British fleet was sent to defend the Falkland the human error? It was at this differences. However, there are no point that the Argentines released their missile models there or in the AI literature of errors and sank an unsuspecting British ship. The operator had The errors result from proficient task performers no inkling of the ramifications of the system practicing in a natural environment; they designs under the current operating conditions. New error and critiquing models operator has virtually no way out. The remarkable need to capture and reflect this difference. We computer-aided design (ICAD) to mitigate begin by examining the design process and such problems. Specifically, we examine the the cognitive difficulties it poses. The designer uses a interference problems are also increasingly variety of cognitive operators to generate a evident on civilian automobiles, airplanes, design, test it under various conditions, refine and ships that cram telephones, radios, computers, it until a stopping rule is reached, and then radar devices, and other electromagnetically store the design as a prototype or analog to incompatible devices into close help start a new process for the next design proximity. The design process is sufficiently complex domain are relevant to all engineering design that a correct and complete design applications that must factor any operational simply cannot be deduced from starting conditions (or manufacturability, sales, or other downstream) or simulation model results.

Architecture) provides a generic domain independent computational architecture suitable for command, planning and execution applications.

In this paper we show that discrete affine wavelet transforms can provide a tool for the analysis and synthesis of standard feedforward neural networks. It is shown that wavelet frames for L2(IR) can be constructed based upon sigmoids. The spatia-spectral localization property of wavelets can be exploited in defining the topology and determining the weights of a feedforward network. Training a network constructed using the synthesis procedure described here involves minimization of a convex cost functional and therefore avoids pitfalls inherent in standard backpropagation algorithms. Extension of these methods to L2(IRN) is also discussed.

Viewing n-variable boolean functions as vectors in'R'2", we invoke tools from linear algebra and linear programming to derive new results on the realizability of boolean functions using threshold gat.es. Using this approach, one can obtain: (1) upper-bounds on the number of spurious memories in HopfielJ networks, and on the number of functions implementable by a depth-d threshold circuit; (2) a lower bound on the number of ort.hogonal input.

Inspired by the information theoretic idea of minimum description length, we add a term to the back propagation cost function that penalizes network complexity. We give the details of the procedure, called weight-elimination, describe its dynamics, and clarify the meaning of the parameters involved. From a Bayesian perspective, the complexity term can be usefully interpreted as an assumption about prior distribution of the weights. We use this procedure to predict the sunspot time series and the notoriously noisy series of currency exchange rates. 1 INTRODUCTION Learning procedures for connectionist networks are essentially statistical devices for performing inductive inference. There is a tradeoff between two goals: on the one hand, we want such devices to be as general as possible so that they are able to learn a broad range of problems.

Sherif M. Botros Christopher G. Atkeson Brain and Cognitive Sciences Department and the Artificial Intelligence Laboratory Massachusetts Institute of Technology Cambridge, MA 02139 Abstract We examine the ability of radial basis functions (RBFs) to generalize. We compare the performance of several types of RBFs. We use the inverse dynamics of an idealized two-joint arm as a test case. We find that without a proper choice of a norm for the inputs, RBFs have poor generalization properties. A simple global scaling of the input variables greatly improves performance.

Spoken input to computers, however, has yet to pass the threshold of practicality. Despite some recent successful demonstrations, current speech recognition systems typically fall far short of human capabilities of continuous speech recognition with essentially unrestricted vocabulary and speakers, under adverse acoustic environments.

Previous work has shown the ability of Multilayer Perceptrons (MLPs) to estimate emission probabilities for Hidden Markov Models (HMMs). The advantages of a speech recognition system incorporating both MLPs and HMMs are the best discrimination and the ability to incorporate multiple sources of evidence (features, temporal context) without restrictive assumptions of distributions or statistical independence. This paper presents results on the speaker-dependent portion of DARPA's English language Resource Management database. Results support the previously reported utility of MLP probability estimation for continuous speech recognition. An additional approach we are pursuing is to use MLPs as nonlinear predictors for autoregressive HMMs. While this is shown to be more compatible with the HMM formalism, it still suffers from several limitations. This approach is generalized to take account of time correlation between successive observations, without any restrictive assumptions about the driving noise. 1 INTRODUCTION We have been working on continuous speech recognition using moderately large vocabularies (1000 words) [1,2].

The development of learning algorithms is generally based upon the minimization of an energy function. It is a fundamental requirement to compute the gradient of this energy function with respect to the various parameters of the neural architecture, e.g., synaptic weights, neural gain,etc. In principle, this requires solving a system of nonlinear equations for each parameter of the model, which is computationally very expensive. A new methodology for neural learning of time-dependent nonlinear mappings is presented. It exploits the concept of adjoint operators to enable a fast global computation of the network's response to perturbations in all the systems parameters. The importance of the time boundary conditions of the adjoint functions is discussed. An algorithm is presented in which the adjoint sensitivity equations are solved simultaneously (Le., forward in time) along with the nonlinear dynamics of the neural networks. This methodology makes real-time applications and hardware implementation of temporal learning feasible.

Viewing n-variable boolean functions as vectors in'R'2", we invoke tools from linear algebra and linear programming to derive new results on the realizability of boolean functions using threshold gat.es. Using this approach, one can obtain: (1) upper-bounds on the number of spurious memories in HopfielJ networks, and on the number of functions implementable by a depth-d threshold circuit; (2) a lower bound on the number of ort.hogonal input.