An automatic iterative unsupervised data analysis tool is presented as a modification of well known Isodata algorithm. The main feature is its complete blindness and repeatability of the obtained results. It automatically selects a suitable number of features able to describe the whole data set requiring only one input parameter. As an application, color vector quantization has been addressed, both on real and on synthetic data sets, showing good performances
IDA - iterative data analysis applied to color vector quantization / D'Orazio, T; Guaragnella, C. - STAMPA. - (2004), pp. 107-110. (Intervento presentato al convegno First International Symposium on Control, Communications and Signal Processing, ISCCSP 2004 tenutosi a Hammamet, Tunisia nel March, 21-24, 2004) [10.1109/ISCCSP.2004.1296230].
IDA - iterative data analysis applied to color vector quantization
Guaragnella C
2004-01-01
Abstract
An automatic iterative unsupervised data analysis tool is presented as a modification of well known Isodata algorithm. The main feature is its complete blindness and repeatability of the obtained results. It automatically selects a suitable number of features able to describe the whole data set requiring only one input parameter. As an application, color vector quantization has been addressed, both on real and on synthetic data sets, showing good performancesI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
