多元异质性视角下的科学基金资助策略优化研究—以人工智能领域为例

Optimization of Science Foundation Funding Strategies from the Perspective of Multiple Heterogeneity —— A Case Study of Artificial Intelligence Field

摘要:

科技项目经费的投入日益增多,因此对于政府资助效果和资助优化的研究越发重要。本文以中国自然科学基金(NSFC)中的人工智能领域为例,构建“学者属性-主题特征-资助模式”的多元异质性框架,运用负二项回归模型、多目标组合和带有精英策略的非支配排序遗传算法(NSGA-Ⅱ)等进行测算。本文识别了多元异质性影响要素及其影响机制,探索了使得科研产出绩效达到最大值的最优解集,进而探究了人才培养导向-前沿引领导向的差异化目标选择的最优策略。本文为科研管理政策的制定与资助政策的改进策略提供量化性、可预见性的参考,从而推动科技进步。

Abstract:

With the increasing investment in science and technology projects, research on the effectiveness and optimization of government funding is becoming more and more important. In this paper, we take the field of artificial intelligence in the Natural Science Foundation of China (NSFC) as an example, and construct a multivariate heterogeneity framework of “scholar attributes-topic characteristics-funding modes”, and measure it by using the negative binomial regression model, multi-objective combinations, and non-dominated sorting genetic algorithms with elite strategies (NSGA-II). This paper identifies the influencing elements of multivariate heterogeneity and their influencing mechanisms, explores the optimal solution set that maximizes the performance of research output, and then explores the optimal strategy for talent cultivation-oriented-frontier-led differentiated goal selection. This paper provides quantitative and predictable references for the formulation of research management policies and the improvement strategies of funding policies, so as to promote scientific and technological progress.