Industry
Balancing Between Bagging and Bumping
We compare different methods to combine predictions from neural networks trained on different bootstrap samples of a regression problem. One of these methods, introduced in [6] and which we here call balancing, is based on the analysis of the ensemble generalization error into an ambiguity term and a term incorporating generalization performances of individual networks. We show how to estimate these individual errors from the residuals on validation patterns. Weighting factors for the different networks follow from a quadratic programming problem. On a real-world problem concerning the prediction of sales figures and on the well-known Boston housing data set, balancing clearly outperforms other recently proposed alternatives as bagging [1] and bumping [8]. 1 EARLY STOPPING AND BOOTSTRAPPING Stopped training is a popular strategy to prevent overfitting in neural networks.
Online Learning from Finite Training Sets: An Analytical Case Study
By an extension of statistical mechanics methods, we obtain exact results for the time-dependent generalization error of a linear network with a large number of weights N. We find, for example, that for small training sets of size p N, larger learning rates can be used without compromising asymptotic generalization performance or convergence speed. Encouragingly, for optimal settings of TJ (and, less importantly, weight decay,\) at given final learning time, the generalization performance of online learning is essentially as good as that of offline learning.
Learning from Demonstration
By now it is widely accepted that learning a task from scratch, i.e., without any prior knowledge, is a daunting undertaking. Humans, however, rarely attempt to learn from scratch. They extract initial biases as well as strategies how to approach a learning problem from instructions and/or demonstrations of other humans. For learning control, this paper investigates how learning from demonstration can be applied in the context of reinforcement learning. We consider priming the Q-function, the value function, the policy, and the model of the task dynamics as possible areas where demonstrations can speed up learning. In general nonlinear learning problems, only model-based reinforcement learning shows significant speedup after a demonstration, while in the special case of linear quadratic regulator (LQR) problems, all methods profit from the demonstration. In an implementation of pole balancing on a complex anthropomorphic robot arm, we demonstrate that, when facing the complexities of real signal processing, model-based reinforcement learning offers the most robustness for LQR problems. Using the suggested methods, the robot learns pole balancing in just a single trial after a 30 second long demonstration of the human instructor.
Rapid Visual Processing using Spike Asynchrony
Thorpe, Simon J., Gautrais, Jacques
We have investigated the possibility that rapid processing in the visual system could be achieved by using the order of firing in different neurones as a code, rather than more conventional firing rate schemes. Using SPIKENET, a neural net simulator based on integrate-and-fire neurones and in which neurones in the input layer function as analogto-delay converters,we have modeled the initial stages of visual processing. Initial results are extremely promising. Even with activity in retinal output cells limited to one spike per neuron per image (effectively ruling out any form of rate coding), sophisticated processing based on asynchronous activation was nonetheless possible.
Learning Exact Patterns of Quasi-synchronization among Spiking Neurons from Data on Multi-unit Recordings
Martignon, Laura, Laskey, Kathryn B., Deco, Gustavo, Vaadia, Eilon
This paper develops arguments for a family of temporal log-linear models to represent spatiotemporal correlations among the spiking events in a group of neurons. The models can represent not just pairwise correlations but also correlations of higher order. Methods are discussed for inferring the existence or absence of correlations and estimating their strength. A frequentist and a Bayesian approach to correlation detection are compared.
Salient Contour Extraction by Temporal Binding in a Cortically-based Network
Yen, Shih-Cheng, Finkel, Leif H.
It has been suggested that long-range intrinsic connections in striate cortex may play a role in contour extraction (Gilbert et aI., 1996). A number of recent physiological and psychophysical studies have examined the possible role of long range connections in the modulation of contrast detection thresholds (Polat and Sagi, 1993,1994; Kapadia et aI., 1995; Kovacs and Julesz, 1994) and various pre-attentive detection tasks (Kovacs and Julesz, 1993; Field et aI., 1993). We have developed a network architecture based on the anatomical connectivity of striate cortex, as well as the temporal dynamics of neuronal processing, that is able to reproduce the observed experimental results. The network has been tested on real images and has applications in terms of identifying salient contours in automatic image processing systems. 1 INTRODUCTION Vision is an active process, and one of the earliest, preattentive actions in visual processing is the identification of the salient contours in a scene. We propose that this process depends upon two properties of striate cortex: the pattern of horizontal connections between orientation columns, and temporal synchronization of cell responses. In particular, we propose that perceptual salience is directly related to the degree of cell synchronization. We present results of network simulations that account for recent physiological and psychophysical "pop-out" experiments, and which successfully extract salient contours from real images.
Adaptive On-line Learning in Changing Environments
Murata, Noboru, Mรผller, Klaus-Robert, Ziehe, Andreas, Amari, Shun-ichi
An adaptive online algorithm extending the learning of learning idea is proposed and theoretically motivated. Relying only on gradient flow information it can be applied to learning continuous functions or distributions, even when no explicit loss function is given and the Hessian is not available. Its efficiency is demonstrated for a non-stationary blind separation task of acoustic signals.
Learning Appearance Based Models: Mixtures of Second Moment Experts
Bregler, Christoph, Malik, Jitendra
This paper describes a new technique for object recognition based on learning appearance models. The image is decomposed into local regions which are described by a new texture representation called "Generalized Second Moments" that are derived from the output of multiscale, multiorientation filter banks. Class-characteristic local texture features and their global composition is learned by a hierarchical mixture of experts architecture (Jordan & Jacobs). The technique is applied to a vehicle database consisting of 5 general car categories (Sedan, Van with backdoors, Van without backdoors, old Sedan, and Volkswagen Bug). This is a difficult problem with considerable in-class variation.
Removing Noise in On-Line Search using Adaptive Batch Sizes
Stochastic (online) learning can be faster than batch learning. However, at late times, the learning rate must be annealed to remove the noise present in the stochastic weight updates. In this annealing phase, the convergence rate (in mean square) is at best proportional to l/T where T is the number of input presentations. An alternative is to increase the batch size to remove the noise. In this paper we explore convergence for LMS using 1) small but fixed batch sizes and 2) an adaptive batch size. We show that the best adaptive batch schedule is exponential and has a rate of convergence which is the same as for annealing, Le., at best proportional to l/T.
Selective Integration: A Model for Disparity Estimation
Gray, Michael S., Pouget, Alexandre, Zemel, Richard S., Nowlan, Steven J., Sejnowski, Terrence J.
Local disparity information is often sparse and noisy, which creates two conflicting demands when estimating disparity in an image region: the need to spatially average to get an accurate estimate, and the problem of not averaging over discontinuities. We have developed a network model of disparity estimation based on disparityselective neurons, such as those found in the early stages of processing in visual cortex. The model can accurately estimate multiple disparities in a region, which may be caused by transparency or occlusion, in real images and random-dot stereograms. The use of a selection mechanism to selectively integrate reliable local disparity estimates results in superior performance compared to standard back-propagation and cross-correlation approaches. In addition, the representations learned with this selection mechanism are consistent with recent neurophysiological results of von der Heydt, Zhou, Friedman, and Poggio [8] for cells in cortical visual area V2. Combining multi-scale biologically-plausible image processing with the power of the mixture-of-experts learning algorithm represents a promising approach that yields both high performance and new insights into visual system function.