Title :
About Neighborhood Counting Measure Metric and Minimum Risk Metric
Author :
Argentini, A. ; Blanzieri, Enrico
Author_Institution :
Dipt. di Ing. e Scienza dell´Inf., Univ. di Trento, Trento, Italy
fDate :
4/1/2010 12:00:00 AM
Abstract :
In a 2006 TPAMI paper, Wang proposed the neighborhood counting measure, a similarity measure for the k-NN algorithm. In his paper, Wang mentioned the minimum risk metric (MRM,), an early distance measure based on the minimization of the risk of misclassification. Wang did not compare NCM to MRM because of its allegedly excessive computational load. In this comment paper, we complete the comparison that was missing in Wang´s paper and, from our empirical evaluation, we show that MRM outperforms NCM and that its running time is not prohibitive as Wang suggested.
Keywords :
learning (artificial intelligence); pattern classification; early distance measure; k-NN algorithm; k-nearest neighbor algorithm; machine learning; minimum risk metric; neighborhood counting measure metric; pattern recognition; risk minimization; MRM; NCM.; Pattern recognition; distance measures; k-Nearest Neighbors; machine learning;
Journal_Title :
Pattern Analysis and Machine Intelligence, IEEE Transactions on
DOI :
10.1109/TPAMI.2009.69