Feature construction is of vital importance in reinforcement learning, as the quality of a value function or policy is largely determined by the corresponding features. The recent successes of deep reinforcement learning (RL) only increase the importance of understanding feature construction. Typical deep RL approaches use a linear output layer, which means that deep RL can be interpreted as a ...

Phoneme classification is investigated for linear feature domains with the aim of improving robustness to additive noise. In linear feature domains noise adaptation is exact, potentially leading to more accurate classification than representations involving non-linear processing and dimensionality reduction. A generative framework is developed for isolated phoneme classification using linear fe...

This paper presents and evaluates two linear feature extractors based on mutual information. These feature extractors consider general dependencies between features and class labels, as opposed to well known linear methods such as PCA which does not consider class labels and LDA, which uses only simple low order dependencies. As evidenced by several simulations on high dimensional data sets, th...

Vectorcardiography is an important area of human and machine pattern recognition. The wide range of interclass variation in observed VCGs is attributed to variations in body structure. IntraManuscript received November 30, 1970; revised April 2, 1971. This research was conducted at Carnegie-Mellon University, Pittsburgh, Pa., and was supported by the National Institute of General Medical Scienc...

Mapping techniques have been regularly used for visualization of high-dimensional data sets. In this paper, mapping to d 2 is studied, with the purpose of feature extraction. Two di erent non-linear techniques are studied: self-organizing maps and auto-associative feedforward networks. The non-linear techniques are compared to linear Principal Component Analysis (PCA). A comparison with respect...

In this work, two new techniques for non-linear feature extraction are presented. In these techniques, new features are obtained as radial projections of the original measurements. Radial projections are a particular kind of second order transformations that show interesting properties: they capture the local structure of the data and reduce dramatically the number of parameters to estimate fro...

We investigate unsupervised pre-training of deep architectures as feature generators for “shallow” classifiers. Stacked Denoising Autoencoders (SdA) [23], when used as feature pre-processing tools for SVM classification, can lead to significant improvements in accuracy – however, at the price of a substantial increase in computational cost. In this paper we create a simple algorithm which mimic...

With microarray gene-expression data, we compare supervised feature extraction methods with the unsupervised feature extraction methods. From experimental results, it is shown that the supervised feature extraction methods are more powerful than the unsupervised feature extraction methods in terms of class separability.

This paper derives a novel linear position constraint for cameras seeing a common scene point, which leads to a direct linear method for global camera translation estimation. Unlike previous solutions, this method deals with collinear camera motion and weak image association at the same time. The final linear formulation does not involve the coordinates of scene points, which makes it efficient...

We describe the dynamics of learning in unsupervised linear feature-discovery networks that have recurrent lateral connections. Bifurcation theory provides a description of the location of multiple equilibria and limit cycles in the weight-space dynamics.