On the mean accuracy of statistical pattern recognizers

  title={On the mean accuracy of statistical pattern recognizers},
  author={Gordon F. Hughes},
  journal={IEEE Trans. Inf. Theory},
  • G. Hughes
  • Published 1968
  • Mathematics
  • IEEE Trans. Inf. Theory
The overall mean recognition probability (mean accuracy) of a pattern classifier is calculated and numerically plotted as a function of the pattern measurement complexity n and design data set size m . Utilized is the well-known probabilistic model of a two-class, discrete-measurement pattern environment (no Gaussian or statistical independence assumptions are made). The minimum-error recognition rule (Bayes) is used, with the unknown pattern environment probabilities estimated from the data… 

Figures and Tables from this paper

The mean accuracy of pattern recognizers with many pattern classes (Corresp.)

  • R. Kain
  • Computer Science, Mathematics
    IEEE Trans. Inf. Theory
  • 1969
Some curves relating the mean performance of a pattern classifier averaged over all pattern recognition problems of given complexity provide a lower bound to the performance of recognizers with more than two pattern classes.

Considerations of sample and feature size

The design-set error rate for a two-class problem with multivariate normal distributions is derived as a function of the sample size per class (N) and dimensionality (L) and is demonstrated to be an extremely biased estimate of either the Bayes or test- set error rate.

Independence of measurements and the mean recognition accuracy

In this paper the following result of practical importance is derived: under certain conditions, having just one sample from class 1, and none at all from class 2, can result in a recognition accuracy arbitrarily close to unity for a large enough number of measurements.

39 Dimensionality and sample size considerations in pattern recognition practice

Effect of dimensionality and estimation on the performance of gaussian classifiers

Exact performance of error estimators for discrete classifiers

Statistical discrimination using inaccurate models

  • D. Kazakos
  • Computer Science
    IEEE Trans. Inf. Theory
  • 1982
The performance of a multiclass maximum likelihood decision rule is analyzed, when inaccurate versions of the true probability density functions are used. A general bound to the error probability is



Statistical Recognition Functions and the Design of Pattern Recognizers

In practice, a prohibitively large sample is required to provide a basis for estimating the optimal recognition function, and certain assumptions about the probability distributions of the measurements are warranted, recognition functions that are asymptotically optimal may be obtained readily.

Pattern recognition preprocessing by similarity functionals

It is shown that design and use of recognizers on such preprocessed pattern data results in increased overall recognition accuracy and reduced storage requirements, and by a reduction of several orders of magnitude in the storage requirements of a pattern recognizer operating on the reduced data, without degrading recognition accuracy.

The characteristic selection problem in recognition systems

This paper examines the notion of a single number statistic for each characteristic which would have certain desirable properties related to the "goodness" of the characteristic, and shows that, in general, no such statistic exists.

Nearest neighbor pattern classification

The nearest neighbor decision rule assigns to an unclassified sample point the classification of the nearest of a set of previously classified points, so it may be said that half the classification information in an infinite sample set is contained in the nearest neighbor.

An optimum character recognition system using decision functions

  • C. Chow
  • Computer Science
    IRE Trans. Electron. Comput.
  • 1957
The character recognition problem, usually resulting from characters being corrupted by printing deterioration and/or inherent noise of the devices, is considered from the viewpoint of statistical

On the effectiveness of receptors in recognition systems

Some of the theoretical problems encountered in trying to determine a more formal measure of the effectiveness of a set of tests are discussed; a measure which might be a practical substitute for the empirical evaluation.

Statistical prediction by discriminant analysis

The limited amount of information contained in a set of meteorological predictors precludes any precise statement concerning which one of a number of possible future events will occur. For purposes

An introduction to the theory of random signals and noise

The aim of this book is to clarify the role of noise in the development of linear and nonlinear systems and to provide a procedure forormalising the noise generated by these systems.

A Selected Descriptor-Indexed Bibliography to the Literature on Artificial Intelligence

This listing is intended as an introduction to the literature on Artificial Intelligence?i.e., to the literature dealing with the problem of making machines behave intelligently. We have divided this

Statistical Communication Theory

This report describes three completed research projects and reports on two current research projects that were completed and two that are in the process of being completed.