构造(python库)
一致性(知识库)
计算机科学
一般化
大数据
人工神经网络
模糊逻辑
相关性(法律)
索引(排版)
工业工程
人工智能
数据挖掘
数学
工程类
万维网
数学分析
程序设计语言
法学
政治学
标识
DOI:10.2478/amns.2023.2.01361
摘要
Abstract This paper analyzes the three-level inclusion relationship of high-level innovative talents and combs the structure of high-level scientific and technological talent evaluation models based on big data technology. Aiming at the evaluation problems of high-level scientific and technological talents, a fuzzy neural network model is constructed, and at the same time, the R&D middle school effect is utilized to evaluate the innovation achievements of high-level scientific and technological talents. Construct the evaluation index system of high-level scientific and technological innovative talents by utilizing 6 first-level indexes, 14 second-level indexes and 48 third-level indexes. Create a hierarchical analysis structure model, evaluate the indicator data through a judgment matrix and consistency test, and output the indicator weights. Analyze the relevance of the indicator model for different input layer neurons in fuzzy hierarchical analysis through comparative experiments. Use empirical analysis to analyze the innovative evaluation scores of high-level scientific and technological talents in Group A. The experimental results show that when the input layer contains 48 neurons, the loss value ranges from [0.132,1.765], the loss decreases the fastest, the stronger the indicator correlation, the stronger the generalization ability of the fuzzy neural network regression model. The overall scores of the evaluation of high-level scientific and technological talents of Group A for the first and second-level indicators are 3.54 and 3.869, respectively, and the overall view of Group A’s high-level scientific and technological talent innovative ability is better. Good.
科研通智能强力驱动
Strongly Powered by AbleSci AI