Recurrence quantification analysis of GSR signals using a neural network recognized public speaking anxiety with average correct rates of 86.70% (calm vs high anxiety) and 78.83% (high vs low anxiety).
Can recurrence quantification analysis of GSR signals accurately recognize public speaking anxiety?
Recurrence quantification analysis of GSR signals combined with a neural network can effectively classify different states of public speaking anxiety.
The aim of this article is to recognize public speaking anxiety based on galvanic skin response through recurrence plot and recurrence quantification analysis. Twenty-two female subjects have participated in this study, and the anxiety state is induced by anticipated and real public speaking. Two nonlinear features, recurrence rate and entropy of diagonal length, are extracted from recurrence plots as the feature set. Furthermore, the paper applies back-propagation neural network algorithm to achieve the goal of binary-classifications between the calmness state and the high anxiety state, as well as between the high anxiety state and the low anxiety state. Experimental results confirm the validity and high performance of this novel method, with an average correct rate of 86.70% and 78.833%, respectively.
Zhang et al. (Sun,) conducted a other in Public speaking anxiety (n=22). Recurrence quantification analysis of GSR signals was evaluated on Average correct classification rate between calmness and high anxiety states, and between high and low anxiety states. Recurrence quantification analysis of GSR signals using a neural network recognized public speaking anxiety with average correct rates of 86.70% (calm vs high anxiety) and 78.83% (high vs low anxiety).
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: