Plurality voting-based multiple classifier systems: statistically independent with respect to dependent classifier sets

Date

2002-11-01

Author

Demirekler, Mübeccel
Altincay, H

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

95
views

0
downloads

The simultaneous use of multiple classifiers has been shown to provide performance improvement in classification problems. The selection of an optimal set of classifiers is an important part of multiple classifier systems and the independence of classifier outputs is generally considered to be an advantage for obtaining better multiple classifier systems. In this paper, the need for the classifier independence is interrogated from classification performance point of view. The performance achieved with the use of classifiers having independent joint distributions is compared to some other classifiers which are defined to have best and worst joint distributions. These distributions are obtained by formulating the combination operation as an optimization problem. The analysis revealed several important observations about classifier selection which are then used to analyze the problem of selecting an additional classifier to be used with the available multiple classifier system. (C) 2002 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.

Subject Keywords

Multiple Classifier Systems , Statistical Classifier Combination , Statistical Pattern Recognition, Classifier Selection , Independent Distributions , Best Distributions , Worst Distributions , Adding New Classifiers , Plurality Voting , Bayesian Formalism

URI

https://hdl.handle.net/11511/56682

Journal

PATTERN RECOGNITION

DOI

https://doi.org/10.1016/s0031-3203(01)00227-8

Collections

Graduate School of Natural and Applied Sciences, Article

Suggestions

OpenMETU
Core

Numerical implementation of magneto-acousto-electrical tomography (MAET) using a linear phased array transducer GÖZÜ, Mehmet Soner; ZENGİN, Reyhan; Gençer, Nevzat Güneri (2018-02-01) In this study, the performance and implementation of magneto-acousto-electrical tomography (MAET) is investigated using a linear phased array (LPA) transducer. The goal of MAET is to image the conductivity distribution in biological bodies. It uses the interaction between ultrasound and a static magnetic field to generate velocity current density distribution inside the body. The resultant voltage due to velocity current density is sensed by surface electrodes attached on the body. In this study, the theory...
Performance Comparison of Different Sparse Array Configurations for Ultra-Wideband, Near-field Imaging Applications Cetin, Beyzat Talat; Alatan, Lale (2017-03-24) The aim of this study is to compare the performance of different multiple-input multiple-output (MIMO) array topologies, intended to be used in ultra-wideband (UWB) near-field imaging applications, by using an analysis method that does not include the effects of image reconstruction algorithm. For this purpose, maximum projection method, previously proposed for the analysis of UWB arrays under far-field conditions, is utilized and modified to obtain two way beam patterns of UWB arrays operating in the near-...
Sliding mode control of linearly actuated nonlinear systems Durmaz, Burak; Özgören, Mustafa Kemal; Department of Mechanical Engineering (2009) This study covers the sliding mode control design for a class of nonlinear systems, where the control input affects the state of the system linearly as described by (d/dt)x=A(x)x+B(x)u+d(x). The main streamline of the study is the sliding surface design for the system. Since there is no systematic way of designing sliding surfaces for nonlinear systems, a moving sliding surface is designed such that its parameters are determined in an adaptive manner to cope with the nonlinearities of the system. This adapt...
Exponential stability of periodic solutions of recurrent neural networks with functional dependence on piecewise constant argument Akhmet, Marat; Cengiz, Nur (The Scientific and Technological Research Council of Turkey, 2018-01-01) In this study, we develop a model of recurrent neural networks with functional dependence on piecewise constant argument of generalized type. Using the theoretical results obtained for functional differential equations with piecewise constant argument, we investigate conditions for existence and uniqueness of solutions, bounded solutions, and exponential stability of periodic solutions. We provide conditions based on the parameters of the model.
BenchMetrics: a systematic benchmarking method for binary classification performance metrics Canbek, Gurol; Taşkaya Temizel, Tuğba; SAĞIROĞLU, ŞEREF (2021-08-01) This paper proposes a systematic benchmarking method called BenchMetrics to analyze and compare the robustness of binary classification performance metrics based on the confusion matrix for a crisp classifier. BenchMetrics, introducing new concepts such as meta-metrics (metrics about metrics) and metric space, has been tested on fifteen well-known metrics including balanced accuracy, normalized mutual information, Cohen's Kappa, and Matthews correlation coefficient (MCC), along with two recently proposed me...

Citation Formats

M. Demirekler and H. Altincay, “Plurality voting-based multiple classifier systems: statistically independent with respect to dependent classifier sets,” PATTERN RECOGNITION, pp. 2365–2379, 2002, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/56682.