Title :
Measures of classification accuracy based on fuzzy similarity
Author :
Jäger, Gunther ; Benz, Ursula
Author_Institution :
Deutsches Zentrum fur Luft- und Raumfahrt DLR, Wessling, Germany
fDate :
5/1/2000 12:00:00 AM
Abstract :
For the validation of classifications, different accuracy measures are known, mostly based on the confusion matrix. The authors develop a common basis for these measures and show that fuzzy similarity not only provides an additional possibility of measuring classification accuracy but also generalizes some of the traditional notions to “fuzzy ground truth” and fuzzy classification
Keywords :
geophysical signal processing; geophysical techniques; image classification; remote sensing; terrain mapping; accuracy; confusion matrix; fuzzy classification; fuzzy ground truth; fuzzy similarity; geophysical measurement technique; image classification; land surface; remote sensing; terrain mapping; validation; Error correction; Fuzzy neural networks; Fuzzy set theory; Image resolution; Image sensors; Neural networks; Spatial resolution;
Journal_Title :
Geoscience and Remote Sensing, IEEE Transactions on
