Research and Implementation of AI Talents Cultivation in “Projects-Driven” Teaching Mode from The Perspective of New Quality Productivity

In this paper, the “projects-driven” teaching model of AI talents training is studied and realized from the perspective of new quality productivity. First, construct a multi-dimensional evaluation system based on 11 teaching (performance) indicators such as understanding, enthusiasm, comprehensiveness, efficiency, practicability and innovation and so on; Second, the evaluation system is used to filter and construct project cases: theoretical analysis and derivation, algorithm design and implementation, data collection and preprocessing, code implementation and experiment, and result analysis, etc. Project cases need to embed knowledge of various disciplines, and teaching and learning are also completed through project cases to index knowledge content, forming a “double-engines” of projects embedding knowledge and projects indexing knowledge; Third, use each teaching (performance) indicators of the evaluation system to evaluate the teaching effect and form a “closed-loop” of “projects-driven” teaching mode; Finally, through performance analysis and comparison, we can get: Suppose that under ideal conditions, the traditional teaching mode, because it is difficult to carry out multi-disciplinary and multi-knowledge teaching in a centralized time and place in practice, even in this case, the “projects-driven” teaching mode greatly exceeds the traditional teaching mode in all teaching (performance) indicators, and all the minimum values are larger than the maximum values, about 20% larger. Among all indicators, the highest values are 99% and 80%, the lowest values are 80% and 50%, the highest average values are 92.55% and 70.18%, and the lowest average values are 86.73% and 62.09%, respectively. The required class hours are also greatly reduced, only 57.69% of the traditional teaching mode, and almost half the class hours are saved. All the results proved that the “projects-driven” teaching mode AI talents training method under the new quality productivity perspective studied in this paper has the characteristics of high efficiency, good stability and excellent overall performance and so on.