This study aims to reveal the scientific associations between the motor competence related physical fitness and the medical health status of college students from south China. Two hundred and fourteen college students, including 112 males and 102 females, from 17 provinces were administrated with the Shantou University fitness test battery twice. All the subjects were asked for a medical examination, including a questionnaire and a physical examination. The records were assessed and concluded by three expertise medical practitioners. A machine learning model equipped with a new loss was designed to deal with the soft label issue. Armed with the trained model, we mine and highlight the relationship between the motor competence related physical fitness and the medical health status of the college students. The physical educators, the educational authorities, the universities, and the individuals may potentially benefit from this study.

Physical test (PF) is demonstrated by a variety of factors including body weight status, cardiorespiratory fitness, musculoskeletal fitness (muscular strength and endurance) and flexibility, which are related to college student's physical condition. The test helps them require an understanding of behavioral attributes and causative mechanisms that promote their physical fitness. Judging the level of college student's physical fitness quickly is very necessary. In this paper, we present an enhanced learning framework for the classification of college student physical fitness. To achieve this, we first design a selection strategy for the students by considering the features that have the greatest impact on. Then, we present an enhanced learning framework equipped with our custom loss function and boosting strategy to classify the level of students' physical fitness. Extensive experiments have demonstrated the outpeformance of our framework.

The technical and emotional control of the javelin game plays an important role for the athlete's performance. However, previous researches show that many psychological methods do not consider the psychological state of the athletes, and take the corresponding measures for psychological adjustment. In this paper, we propose an improved Inverted-U-Type hypothesis method for athlete's psychological adjustment during the training time before competition. According to the athletes' psychological state, adjusting the factors, we carry the weak and strong interventions to adjust psychological and physiological state of the athletes to an optimal level. The results show that the adjustment of the athletes' psychological by our method can effectively eliminate the impact of internal and external factors to improve the stability, which is also help for the athletes' psychological training of other sports.

Sparse model is widely used in hyperspectral image classification. However, different of sparsity and regularization parameters has great influence on the classification results. In this paper, a novel adaptive sparse deep network based on deep architecture is proposed, which can construct the optimal sparse representation and regularization parameters by deep network. Firstly, a data flow graph is designed to represent each update iteration based on Alternating Direction Method of Multipliers (ADMM) algorithm. Forward network and Back-Propagation network are deduced. All parameters are updated by gradient descent in Back-Propagation. Then we proposed an Adaptive Sparse Deep Network. Comparing with several traditional classifiers or other algorithm for sparse model, experiment results indicate that our method achieves great improvement in HSI classification.

No abstract