Abstract-The evaluation of physical education (PE) multimedia teaching refers to the prediction of physical education multimedia teaching quality in the absence of initial multimedia teaching information. Therefore, the evaluation method of PE multimedia teaching based on unsupervised feature learning has achieved good performance. But, its quality prediction accuracy decreases significantly with the reduction of the feature dimension. In order to overcome this defect, the author combines the active learning strategy with the unsupervised feature learning and proposes a kind of data assimilation framework to improve the discriminability of the representation of teaching features. The results show that the proposed method can enhance the accuracy of teaching quality prediction by 8%. Experiments show that, when feature dimension is relatively low, the proposed method can improve the teaching quality prediction accuracy by 8% compared with the method based on unsupervised feature learning. At the same time, the performance of the proposed method is superior to that of the other physical education multimedia teaching evaluation methods at present.Keywords-active learning; dictionary learning; feature learning; teaching presentation; teaching quality evaluation
IntroductionWith the popularization of teaching acquisition equipment such as mobile phones and cameras and the rapid development of social networks, teaching plays an increasingly important role in our daily life. However, due to the limitation of the performance of the teaching processing system, various kinds of distortion will be introduced in the process of teaching access, transmission and processing. The introduction of distortion will reduce the aesthetics of teaching, which, at the same time, will impeding people from obtaining information from the teaching. Therefore, it is necessary to study the teaching quality evaluation methods, so as to apply them to the monitoring of the teaching acquisition equipment performance, the optimal selection of teaching processing system and the optimization of teaching methods.As people are the ultimate recipients of teaching, the subjective evaluation of the quality of teaching by people is the most reliable method [1]. However, the subjective quality evaluation method is time consuming, labor intensive and difficult to be inte-
30