Development Of Using Mobile Device For Fatigue Detection
The main purpose of this research is to estimate prevalence rate of fatigue and construct a fatigue predictive model. By using mobile device, the physiological parameters are collected and measured precisely. The results show that the first 3 days of the fatigue prevalence of graduate students was 76.7%, the last 3 days of the fatigue prevalence of graduate students was 43.3%. Fatigue prevalence rate has decreased significantly, Indicating that the contents of the reminding mechanism does help to reduce the occurrence of fatigue, the reminding is also good. In addition to collect background of life changes, the physiological parameters are collected and measured precisely and CIS score, as the inverted transmission of the Neural Network input and output values. Finally, establish predictive model by Back Propagation Neural Network Model of the accuracy was 87.5%, the sensitivity was 97.56%; the specificity was 74.2%; and AUC was 0.841. Through predictive model to verify the accuracy, the determination of the degree of fatigue has a better way.
Keywords - Fatigue, Mobile Device, Graduate Students, Neural Network, Predictive Model