Construction of Scientific Talents Training Evaluation System Based on Improved Forest Random Algorithm
Wei Ma, Yongjun Qi · 2022
The cultivation of scientific talents is an urgent need to achieve national prosperity, national rejuvenation and people's happiness. At present, it has received extensive attention and thinking from all walks of life, and has become the mainstream value orientation for the purpose of talent cultivation in higher education in my country. It fully reflects the needs of the times and social responsibility to crack this issue. This paper first uses the demand dimension model to analyze the differences between the perspectives of teachers and students in talent training in colleges and universities, which is conducive to the unification of training models and structures. This paper establishes a random forest classification architecture based on the C4.5 algorithm, calculates the information gain rate of different statistical features, selects the statistical feature with prominent information gain rate as the root node to construct multiple decision trees, and composes the generated multiple trees into a random forest. And adopt the simple majority voting mechanism to obtain the classification results. The cultivation of scientific talents is a social system project, involving many complex historical reasons and social factors. It is necessary to pay attention to the institutional mechanisms and cultural traditions at the macro level of the country and society, and to analyze the university spirit and training mode at the meso level of colleges and universities. The final results of the research show that when the maximum tree depth on the dataset is 3dm, the accuracy rates of ID3 algorithm, RF algorithm and DPRF algorithm are 84.56%, 80.66% and 86.89%, respectively, and the improved forest random DPRF algorithm has the highest accuracy. It has certain feasibility in the construction of scientific personnel training evaluation system.