Abstract: In order to scientifically evaluate talent incubating efficiency of technology enterprise incubators, a technology enterprise talent incubating efficiency evaluation model based on S-FNN is proposed. A scientific enterprise incubating efficiency evaluation index system is constructed based on basic process of talent incubating. Fuzzy subtractive clustering method is used to deal with network rules, reduce the rule number of neural network, lest cause network parameters inflation which is hard to train and particle swarm and genetic algorithm is used to determine the fuzzy network parameters. Fuzzy neural network is employed for overall evaluation of technology enterprise incubating efficiency. An actual case demonstrates the feasibility and efficiency of evaluation model.

Key words: technology enterprise incubator, talent incubate, efficiency evaluation, subtractive cluster, fuzzy neural network

摘要： 为科学测评科技企业孵化器的人才孵化效率，提出科技企业人才孵化效率S-FNN评价模型。在研究科技企业人才孵化基本过程的基础上，构建了科技企业人才孵化效率评价指标体系；运用模糊减聚类法对网络规则进行处理，减少神经网络规则数目，以免参数膨胀导致网络难以训练，采用粒子群与遗传算法确定模糊网络参数值；运用模糊神经网络对科技企业人才孵化效率进行综合评价；通过应用案例验证了该评价模型的有效性和可行性。

关键词: 科技企业孵化器, 人才孵化, 效率评价, 减聚类, 模糊神经网络