Table_3.xls (9.5 kB)
Single features derived from pattern recognition paradigms.
dataset
posted on 2013-02-19, 01:29 authored by Jochen Klucken, Jens Barth, Patrick Kugler, Johannes Schlachetzki, Thore Henze, Franz Marxreiter, Zacharias Kohl, Ralph Steidl, Joachim Hornegger, Bjoern Eskofier, Juergen WinklerSelected features from derived from pattern recognition algorithm “APD” that show the highest difference (p<0.00001) if tested for differences between PD patients and controls (Student's T-test). Only low sensitivity and specificity is reached by single feature classification (AdaBoost). Description of feature includes the task and the sensor type/axis.