A sparse kernelized matrix learning vector quantization model for human activity recognition.
Marika Kästner, Marc Strickert, Thomas Villmann · 2013
Abstract. The contribution describes our application to the ESANN'2013 Competition on Human Activity Recognition (HAR) using Android-OS smartphone sensor signals. We applied a kernel variant of learning vector quantization with metric adaptation using only one proto-type vector per class. This sparse model obtains very good accuracies and additionally provides class correlation information. Further, the model allows an optimized class visualization. 1