We present an improvement of Novikoff's perceptron convergence theorem. Reinterpreting this mistake bound as a margin dependent sparsity guarantee allows us to give a PACstyle generalisation error bound for the classifier learned by the perceptron learning algorithm. The bound value crucially depends on the margin a support vector machine would achieve on the same data set using the same kernel. Ironically, the bound yields better guarantees than are currently available for the support vector solution itself.

For SNR-estimation, the input signal is transformed into so-called Amplitude Modulation Spectrograms (AMS), which represent both spectral and temporal characteristics of the respective analysis frame, and which imitate the representation of modulation frequencies in higher stages of the mammalian auditory system. A neural network is used to analyse AMS patterns generated from noisy speech and estimates the local SNR. Noise suppression is achieved by attenuating frequency channels according to their SNR. The noise suppression algorithm is evaluated in speakerindependent digit recognition experiments and compared to noise suppression by Spectral Subtraction. 1 Introduction One of the major problems in automatic speech recognition (ASR) systems is their lack of robustness in noise, which severely degrades their usefulness in many practical applications. Several proposals have been made to increase the robustness of ASR systems, e.g. by model compensation or more noise-robust feature extraction [1, 2]. Another method to increase robustness of ASR systems is to suppress the background noise before feature extraction. Classical approaches for single-channel noise suppression are Spectral Subtraction [3] and related schemes, e.g.

We analyze Gallager codes by employing a simple mean-field approximation that distorts the model geometry and preserves important interactions between sites. The method naturally recovers the probability propagation decoding algorithm as an extremization of a proper free-energy. We find a thermodynamic phase transition that coincides with information theoretical upper-bounds and explain the practical code performance in terms of the free-energy landscape.

Low-dimensional representations are key to solving problems in highlevel vision, such as face compression and recognition. Factorial coding strategies for reducing the redundancy present in natural images on the basis of their second-order statistics have been successful in accounting for both psychophysical and neurophysiological properties of early vision. Class-specific representations are presumably formed later, at the higher-level stages of cortical processing. Here we show that when retinotopic factorial codes are derived for ensembles of natural objects, such as human faces, not only redundancy, but also dimensionality is reduced. We also show that objects are built from parts in a non-Gaussian fashion which allows these local-feature codes to have dimensionalities that are substantially lower than the respective Nyquist sampling rates.

How should we decide among competing explanations of a cognitive process given limited observations? The problem of model selection is at the heart of progress in cognitive science. In this paper, Minimum Description Length (MDL) is introduced as a method for selecting among computational models of cognition. We also show that differential geometry provides an intuitive understanding of what drives model selection in MDL. Finally, adequacy of MDL is demonstrated in two areas of cognitive modeling.

Our al exploits the special structure of the sum of squared error measure in Equation (1); hence, the other objective functions are outside the scope of this paper. The gradient vector and Hessian matrix are given by g g(9) JT rand H H(9) JT J S, where J is the m x n Jacobian matrix of r, and S denotes the matrix of second-derivative terms. If S is simply omitted based on the "small residual" assumption, then the Hessian matrix reduces to the Gauss-Newton model Hessian: i.e., JT J. Furthermore, a family of quasi-Newton methods can be applied to approximate term S alone, leading to the augmented Gauss-Newton model Hessian (see, for example, Mizutani [2] and references therein).

We present a bound on the generalisation error of linear classifiers in terms of a refined margin quantity on the training set. The result is obtained in a PAC-Bayesian framework and is based on geometrical arguments in the space of linear classifiers. The new bound constitutes an exponential improvement of the so far tightest margin bound by Shawe-Taylor et al. [8] and scales logarithmically in the inverse margin. Even in the case of less training examples than input dimensions sufficiently large margins lead to nontrivial bound values and - for maximum margins - to a vanishing complexity term. Furthermore, the classical margin is too coarse a measure for the essential quantity that controls the generalisation error: the volume ratio between the whole hypothesis space and the subset of consistent hypotheses.

In this paper, we discuss some new research directions in automatic speech recognition (ASR), and which somewhat deviate from the usual approaches. More specifically, we will motivate and briefly describe new approaches based on multi-stream and multi/band ASR. These approaches extend the standard hidden Markov model (HMM) based approach by assuming that the different (frequency) channels representing the speech signal are processed by different (independent) "experts", each expert focusing on a different characteristic of the signal, and that the different stream likelihoods (or posteriors) are combined at some (temporal) stage to yield a global recognition output. As a further extension to multi-stream ASR, we will finally introduce a new approach, referred to as HMM2, where the HMM emission probabilities are estimated via state specific feature based HMMs responsible for merging the stream information and modeling their possible correlation.

The resulting combinatorial problem - finding the best agreement half-space over an input sample - is NP hard to approximate to within some constant factor. We suggest a way to circumvent this theoretical bound by introducing a new measure of success for such algorithms. An algorithm is ILmargin successful if the agreement ratio of the half-space it outputs is as good as that of any half-space once training points that are inside the ILmargins of its separating hyper-plane are disregarded. We prove crisp computational complexity results with respect to this success measure: On one hand, for every positive IL, there exist efficient (poly-time) ILmargin successful learning algorithms. On the other hand, we prove that unless P NP, there is no algorithm that runs in time polynomial in the sample size and in 1/ IL that is ILmargin successful for all IL O. 1 Introduction We consider the computational complexity of learning linear perceptrons for arbitrary (Le.

Large margin linear classification methods have been successfully applied to many applications. For a linearly separable problem, it is known that under appropriate assumptions, the expected misclassification error of the computed "optimal hyperplane" approaches zero at a rate proportional to the inverse training sample size. This rate is usually characterized by the margin and the maximum norm of the input data. In this paper, we argue that another quantity, namely the robustness of the input data distribution, also plays an important role in characterizing the convergence behavior of expected misclassification error. Based on this concept of robustness, we show that for a large margin separable linear classification problem, the expected misclassification error may converge exponentially in the number of training sample size.