Edge-Cloud Collaboration Quality Measurement for Physical Education

Abstract

The public physical education is always outdoors, which makes the real-time evaluation of teaching ability difficult for physical education teachers. In order to adapt to the outdoor environment, this paper proposes an edge-cloud collaboration-based physical education evaluation method. First, the wearable devices are worn by students to collect their real-time status. Second, the students' data are transmitted to a cloud server via a wireless network. In the cloud server, a deployed AI model is used to evaluate the physical class. The quality of physical education is divided into five ranks in which there is a strict ordinal relation. In order to reflect the ordinal relation, this paper adopts support vector ordinal regression (SVOR) as the AI model. The SVOR model is learned offline using the students' data from wearable devices and the scores from experts. The scores include teaching attitude, teaching implementation, teaching academia, and teaching development. The simulation shows that the proposed physical education evaluation method can return the real-time quality result. Compared with traditional classification models, the SVOR can achieve much less mean absolute error (MAE) due to considering the ordinal relation in it.