The challenges of measuring, managing, and predicting negative side effects (i.e. cybersickness) during virtual environment (VE) interaction and unwanted aftereffects continue to persist. Most assessments have been limited to subjective questionnaires. This study investigated the effects of VE interaction on nausea (gastric activity) and the development of a quantitative model affording cybersickness prediction. This cybersickness state estimation model (CSEM) offers a quantitative (i.e. physiological based) predictability to cybersickness. The model relates time, recordable gastric activity using electrogastrogram (EGG) measures as a physiological predictor and subjective sickness estimates. This work employs neural network algorithms to model user state associated with cybersickness. Neural network models were designed, trained, and tested to determine the network trainability and accuracy. Mean SSQ total severity scores indicate significant differences between conditions. All CSEM models had a low sum of squared error. The models were found to be accurate in predicting cybersickness from unseen data.
Six theoretical-methodological approaches to analyze written texts in a year-long Blended Learning cours
