Showing papers on "Feature selection published in 1975"

Computations for Variable Selection in Discriminant Analysis

Abstract: An algorithm is proposed for computing statistics for all possible subsets of variables for a discriminant analysis. Optimal subsets of any given size can then be determined. A comparison with a stepwise procedure is also presented through two examples.

Mathematical methods of feature selection in pattern recognition

Abstract: In the 15 years of its existence pattern recognition has made considerable progress on both the theoretical and practical fronts. Starting from the original application of pattern recognition techniques to the problem of character recognition at the time when pattern recognition was conceived these techniques have now penetrated such diverse areas of science as medical diagnosis, remote sensing, finger prints and speech recognition, image classification, etc.* This wide applicability derives from the inherent generality of pattern recognition, which is a direct consequence of the adopted threestage concept of pattern recognition process. According to this concept the process of pattern recognition is viewed as a sequence of three independent functions--representation, feature selection and classification (Fig. 1). Among these functions only the representation stage, which transforms the input patterns into a form suitable for computer processing, is problemdependent. Both the feature selector, the function of which is to reduce the dimensionality of the representation vector, and the classifier, which carries out the actual decision process, work with a vector of measurements which can be considered as an abstract pattern. As a result, the feature selection and classification stages can be implemented using mathematical methods irrespective of the original application. Naturally, this has had a beneficial effect on the progress in the theory of pattern recognition. Although all three stages of the pattern recognition system play an essential role in the process of classifying patterns by machine, the quality of the system's performance depends chiefly on the feature selector. The reasons

On Minimizing the Probability of Misclassification for Linear Feature Selection

Abstract: The use of techniques for feature selection permits treatment of classification problems in spaces of reduced dimensions. A method is considered of linear feature selection for n-dimensional observation vectors which belong to one of two populations, where each population is described by a known multivariate normal density function. More specifically, the problem of finding a 1xn transformation matrix B for which the probability of misclassification with respect to the one-dimensional transformed density functions was minimized was considered. Theoretical results are presented which give rise to a numerically tractable expression for the variation in the probability of misclassification with respect to B. Using this expression a computational procedure is discussed for obtaining a B which minimizes the probability of misclassification. Preliminary numerical results are discussed.

Feature Selection in Pattern Recognition

Abstract: The subject of pattern recognition can be divided into two main areas of study: (1) feature selection and (2) classifier design, as summarized in Fig. 10.1. x(t) is a signal that belongs to K classes denoted by C 1 , C 2 ,..., C K .

A Unified Approach to Feature Selection and Learning in Unsupervised Environments

Abstract: Here the twin problems of feature selection and learning are tackled simultaneously to obtain a unified approach to the problem of pattern recognition in an unsupervised environment. This is achieved by combining a feature selection scheme based on the stochastic learning automata model with an unsupervised learning scheme such as learning with a probabilistic teacher. Test implementation of this scheme using the remotely sensed agricultural data of the Purdue laboratory for agricultural remote sensing (LARS) in a simulated unsupervised mode, has brought out the efficacy of this integrated system of feature selection and learning.

Feature selection for nonlinear stochastic system classification

Abstract: A decision-theoretic formulation is given for the problem of classifying an unknown nonlinear stochastic system into one of M classes when only input-output measurements are available. This leads directly to a pattern recognition solution for the problem, and Bayes-risk theory yields the likelihood-ratio test for class determinations. Parameterizations which yield an implicit description for unknown nonlinear systems are considered, and the theoretical likelihood ratio is related to these parameterizations. The difficult problem of initial feature selection is considered in terms of a parameter vector, and in terms of a quasi-moment expansion, both of which require no a priori knowledge of the system. Experimental results are also cited which show that classification can be accomplished with a low probability of error, and analogies with other classification problems are noted.

Suboptimal method of feature selection and ordering for multiclass problems

Abstract: Feature extraction remains a problem in pattern recognition, and the letter describes a method of weighting features so that the significant parameters of the input patterns are emphasised. This enables dimensionality reduction to take place without loss of information necessary to distinguish between classes.

Optimal selection of passes

Abstract: Preliminary numerical results obtained from the application of a linear feature selection technique to the determination of combinations of passes which best discriminate between a given set of crops in a given area of interest, are reported. The results obtained are not purported to hold in a general situation, but only for the given set of crops and the given, but unknown, levels of several factors-such as soil type, and fertilizer practice, holding in the area of interest. However, by identifying the various factors affecting the spectral signatures, and by formulating a regression model one could use the feature selection technique to determine the regression coefficients for predicting optimal passes for a given set of crops. Another use of the feature selection technique as applied to multiple pass registered data is the generation of enhanced grey scale displays by using a single linear combination of all channels of all designated passes as opposed to a single channel within a single pass.

Feature selection for diagnosis of vectorcardiograms

Abstract: The automatic classification of vectorcardiograms and electrocardiograms into disease classes using computerized pattern recognition techniques has been a much studied problem. To date, however, no system exists which meets desired accuracy and noise immunity requirements and development of new techniques continues. An important aspect of the problem is that of feature selection, in which the functions of data reduction and information preservation are performed. In this paper, the problem of linear feature extraction is studied and a modified form of the Karhunen-Loeve expansion is developed which appears to have some advantages for the present application. Comparison with other feature selection methods is made using a two-dimensional example. Finally, some areas for future research are pointed out.

On the variational equations for Householder transformations in feature selection

Abstract: Results that suggest the possibility of using a sequential monotone process for solving the feature selection problem using Householder transformations are applied to the divergence separability criterion and an expression for the gradient of the divergence with respect to the generator of a single Householder transformation will be developed. This expression for the gradient is used in any number of differential correction schemes (iterators) that attempt to extremize the divergence. Data sets provided by the Earth Observations Division-JSC are used to demonstrate selecting the Householder transformations that generate the kxn matrix defining the best (in the sense of extremizing the divergence) k linear combinations of features. The tests allow initial comparisons to be made with results. In particular, this technique does not appear to require initial guesses for the iterator to be generated without replacement, exhaustive search, or other similar schemes.

Target Spectra Feature Selection Using a Priori Information

Abstract: This correspondence gives a method of selecting n features out of k (n < k) FFT filter outputs which characterize them classes of target signatures to be recognized. The method differs from other known techniques in that it uses a priori information of the target measurements. It is based on an error bound established by Lainiotis which is a weighted sum of Bhattacharyya coefficients.

On designing simulation models for evaluating discriminant analysis routines

Abstract: This paper gives certain simulation models that can be used to verify a Bayes' discriminant algorithm and a feature selection routine foi their accuracy. The verification is developed in terms of statistical tests and the evaluation techniques are suggested, assuming the multivariate normal model for underlying classes. The algorithm evaluation procedures are discussed from a practical viewpoint and are easy to perform.