Title :
A theoretical study on six classifier fusion strategies
Author :
Kuncheva, Ludmila I.
Author_Institution :
Sch. of Informatics, Univ. of Wales, Bangor, UK
fDate :
2/1/2002 12:00:00 AM
Abstract :
We look at a single point in feature space, two classes, and L classifiers estimating the posterior probability for class ω1 . Assuming that the estimates are independent and identically distributed (normal or uniform), we give formulas for the classification error for the following fusion methods: average, minimum, maximum, median, majority vote, and oracle
Keywords :
estimation theory; optimisation; pattern classification; probability; classification error; classifier combination; classifier fusion strategies; feature space; fusion methods; identically distributed estimates; independent classifiers; majority vote; order statistics; posterior probability; theoretical error; Diversity reception; Error analysis; Gaussian distribution; Pattern recognition; Probability; Statistical distributions; Voting;
Journal_Title :
Pattern Analysis and Machine Intelligence, IEEE Transactions on
