Evaluating social anxiety through pulse transit time series based on extreme learning machine
Hui Hu, Guangyuan Liu, Wanhui Wen, Xiaokai Liu · 2016
Social anxiety (SA) would cause substantial distress and functional impairment. People who falls into SA would produce excessive catecholamine, leading to the elevation of arterial blood pressure (ABP). Moreover, pulse transit time (PTT) has a high correlation with ABP. This paper puts forward to use the PTT series to evaluate SA. The subjects' high anxiety state and low anxiety state are respectively induced by impromptu speaking with 21 audience and without audience. The mean of the PTT series is extracted as the effective feature. Two sample t-test between the low anxiety state and the high anxiety state shows that the higher degree of anxiety, the smaller mean of the PTT series. Extreme learning machine and leave-one-out cross validation are used for the recognition and classification. The result shows that the mean training accuracy and mean testing accuracy are 94.82% and 93.17%, respectively.