Skip to main content
. Author manuscript; available in PMC: 2016 Mar 23.
Published in final edited form as: IEEE Trans Hum Mach Syst. 2014 Nov 6;45(5):575–585. doi: 10.1109/THMS.2014.2362529

Fig. 5.

Fig. 5

AUC scores for supervised learning on the scripted and unscripted datasets. AR-DT, AR-LR, AR-BOOST and AR-SVM show significant improvement over the baseline techniques of Fixed and Random in the scripted setting while only AR-DT and AR-LR outperform the baselines in the unscripted setting. Error bars represent the standard error of the mean.