Affect detection from non-stationary physiological data using ensemble classifiers

Affect detection from physiological signals has received considerable attention. One challenge is that physiological measures exhibit considerable variations over time, making classification of future data difficult. The present study addresses this issue by providing insights on how diagnostic physiological features of affect change over time. Affective physiological data (electrocardiogram, electromyogram, skin conductivity, and respiration) was collected from four participants over five sessions each. Classification performance of a number of training strategies, under different conditions of features selection and engineering, were compared using an adaptive classifier ensemble algorithm. Analysis of the performance of individual physiological channels for affect detection is also provided. The key result is that using pooled features set for affect detection is more accurate than using day-specific features. A decision fusion strategy which combines decisions from classifiers trained on individual channels data outperformed a features fusion strategy. Results also show that the performance of the ensemble is affected by the choice of the base classifier and the alpha factor used to update the member classifiers of the ensemble. Finally, the corrugator and zygomatic facial EMGs were found to be more reliable measures for detecting the valence component of affect compared to other channels.