“…Electroencephalogram (EEG)-based emotion classification has led to an emerging and challenging track in affective brain-computer interface (ABCI) domain (Mühl et al, 2014 ; Lin et al, 2015b ). Referring to the prior studies (Chanel et al, 2009 ; Frantzidis et al, 2010 ; Lin et al, 2010b , 2014 , 2015a ; Petrantonakis and Hadjileontiadis, 2010 ; Koelstra et al, 2012 ; Soleymani et al, 2012 ; Hadjidimitriou and Hadjileontiadis, 2013 ; Koelstra and Patras, 2013 ; Jenke et al, 2014 ), a potential bottleneck of developing a practicable emotion-classification model for an individual could be the lack of using sufficient ecologically-valid data, especially for the works (Koelstra et al, 2012 ; Soleymani et al, 2012 ; Koelstra and Patras, 2013 ; Lin et al, 2015a ) using long-duration emotion elicitation, such as movie watching and music listening.…”