Sleep classification with a combination of symbolic learning and learning vector quantization
Gert Pfurtscheller, Doris Flotzinger, Miroslav Kubát · Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society · 1992
Besides statistical methods, various Artificial Intelligence approaches can be used for sleep classification. Learning vector quantization (LVQ) and the top-down induction of decision trees (TDIDT) were applied on 8-hour sleep data from infants. It was shown that with a combination of TDIDT and LVQ the input dimension of the LVQ can be reduced without decreasing the classification accuracy. Classification accuracy was between 67 and 76%, depending on the infant.