To further analyze the content of “dual-teacher” teacher team construction in higher vocational colleges and universities, this paper focuses on teacher moral construction, teacher training, assessment and evaluation, and incentives for in-depth investigation. It is clear that higher vocational colleges and universities should distinguish between “dual-teacher” teacher quality assessment and ordinary teacher assessment, and put forward the gradient “dual-teacher” teacher quality team construction. Improve the support vector machine, use the binary tree to propose an evaluation model based on incomplete DBT-SVM, and combine multiple binary classifiers to solve the multi-classification problem of “dual-teacher” teacher quality evaluation. The optimal parameter combinations, i.e., C =28 and γ = 0.0534, are obtained using the kernel function and parameter tuning experiments, and the accuracy of the model prediction results reaches 94.325%, taking the “dual-teacher” teachers in a university in Y province as the specific evaluation object. This shows that the accuracy of this DBT-SVM-based evaluation method of “dual-teacher” teacher quality is good.