Wrapper Feature Selection for Small Sample Size Data Driven by Complete Error Estimates

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F12%3A00194629" target="_blank" >RIV/68407700:21230/12:00194629 - isvavai.cz</a>

Alternative codes found

RIV/00216208:11110/12:11917 RIV/00064165:_____/12:11917

Result on the web

<a href="http://www.sciencedirect.com/science/article/pii/S0169260712000582" target="_blank" >http://www.sciencedirect.com/science/article/pii/S0169260712000582</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.cmpb.2012.02.006" target="_blank" >10.1016/j.cmpb.2012.02.006</a>

Result language

angličtina

Original language name

Wrapper Feature Selection for Small Sample Size Data Driven by Complete Error Estimates

Original language description

This paper focuses on wrapper-based feature selection for a 1-nearest neighbor classifier. We consider in particular the case of a small sample size with a few hundred instances, which is common in biomedical applications. We propose a technique for calculating the complete bootstrap for a 1-nearest-neighbor classifier (i.e., averaging over all desired test/train partitions of the data). The complete bootstrap and the complete cross-validation error estimate with lower variance are applied as novel selection criteria and are compared with the standard bootstrap and cross-validation in combination with three optimization techniques - sequential forward selection (SFS), binary particle swarm optimization (BPSO) and simplified social impact theory based optimization (SSITO). The experimental comparison based on ten datasets draws the following conclusions: for all three search methods examined here, the complete criteria are a significantly better choice than standard 2-fold cross-validat

Czech name

—

Czech description

—

Type

J_x - Unclassified - Peer-reviewed scientific article (Jimp, Jsc and Jost)

CEP classification

JC - Computer hardware and software

OECD FORD branch

—

Project

<a href="/en/project/GA309%2F09%2F1145" target="_blank" >GA309/09/1145: Mechanisms of deep brain stimulation: Role of the subthalamus in motor, visual and affective processing</a><br>

Continuities

Z - Vyzkumny zamer (s odkazem do CEZ)

Publication year

2012

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Name of the periodical

Computer Methods and Programs in Biomedicine

ISSN

0169-2607

e-ISSN

—

Volume of the periodical

108

Issue of the periodical within the volume

1

Country of publishing house

IE - IRELAND

Number of pages

13

Pages from-to

138-150

UT code for WoS article

000309443000013

EID of the result in the Scopus database

—