Brand is the largest commercial property; its birth is closely related to competition. National sports brand competitiveness is the embodiment of the rise and fall of the sports industry. In order to make up the deficiency of previous empirical research, using the calculation method of PLS-SEM as the machine learning method, mining the sports brand data, and combined with the characteristics of sports industry, to construct the model of sports brand competitiveness, and to promote the development of the sports brand.