基于改进人工鱼群算法和模板驱动的智能培训平台研究

发布时间：2018-05-22 08:36

本文选题：智能培训 + 人工鱼群算法　；参考：《浙江理工大学》2017年硕士论文

【摘要】：随着人工智能和计算机技术的迅速发展,低智能化的手动组卷和随机组卷已经难以满足当代社会的需求。不管从组卷时间上还是组卷质量上都直接或者间接地影响着学习者的学习与进步。为提高组卷效率,合理结合教育测量学理论,本文应用人工鱼群算法和模板驱动技术,构造组卷结构的宏观规划数学模型,建立基于改进人工鱼群算法和模板驱动的培训系统,为智能组卷的实现提供了强有力的支撑。其主要包含以下几个方面:(1)查阅大量国内外培训系统文献,对培训系统的发展以及相关技术进行分析,总结了我国培训系统中存在的一些问题,确定了论文的研究方向。(2)培训系统的理论基础研究。论述了智能培训系统中群智能算法的相关理论,以及Web Service和Nginx、Tomcat的技术支持,确定了培训系统的技术基础。通过人工鱼群模型的建立,增强了试题的搜索能力,简化了目标函数,解决了以往组题中存在的不合理性,确定了智能组卷的方法路线。(3)组卷算法的改进与应用。一般的组卷算法用时多,成功率低,生成的试卷很难满足学员的需要。在采用具有鲁棒性强、全局收敛性好、对初值敏感小的人工鱼群算法后,解决了一般算法的缺陷,但面对多目标问题上还存在不确定性。根据存在的问题,采用拥有吞食行为的全局人工鱼群算法,并与模拟退火算法相结合,最后将改进的人工鱼群算法应用到组卷上并进行测试,证明了该算法组卷的高效性和敏捷性。(4)银行业务页面的变化影响着培训系统题型页面的变化,造成大量的维护和金钱的支出。针对这一问题,提出了模板驱动,增强了出题页面的灵活性。(5)根据智能培训系统的设计目标,进行了详细的设计,并完成了系统的开发。从而验证该系统操作的便捷性以及功能的强大性,使得银行培训更加智能化、高效化、经济化。
[Abstract]:With the rapid development of artificial intelligence and computer technology, it is difficult to meet the needs of modern society with low intelligent manual test paper and random test paper. Both the time and the quality of the test paper directly or indirectly affect the learners' learning and progress. In order to improve the efficiency of composition and reasonably combine the theory of educational measurement, this paper applies artificial fish swarm algorithm and template driven technology, constructs the mathematical model of macro planning of composition structure, and establishes a training system based on improved artificial fish swarm algorithm and template driving. It provides a strong support for the realization of intelligent generating test paper. It mainly includes the following several aspects: 1) consult a large number of domestic and foreign training system literature, analyze the development of training system and related technology, summarize some problems existing in our training system. Determine the research direction of the paper. 2) the theoretical basis of the training system. This paper discusses the related theory of swarm intelligence algorithm in intelligent training system, and the technical support of Web Service and Nginx Tomcat, and determines the technical foundation of the training system. Through the establishment of artificial fish swarm model, the searching ability of test questions is enhanced, the objective function is simplified, the irrationality existing in the former group questions is solved, and the improvement and application of the method route. Due to the low success rate and more time, the test paper is difficult to meet the needs of the students. After adopting artificial fish swarm algorithm with strong robustness, good global convergence and low sensitivity to initial value, the defects of the general algorithm are solved, but there is still uncertainty in the multi-objective problem. According to the existing problems, the global artificial fish swarm algorithm with devouring behavior is adopted and combined with simulated annealing algorithm. Finally, the improved artificial fish swarm algorithm is applied to the test paper and tested. It is proved that the high efficiency and agility of the algorithm can affect the change of the title page of the training system and result in a large amount of maintenance and money expenditure. In order to solve this problem, a template driver is proposed, which enhances the flexibility of the title page. (5) according to the design goal of the intelligent training system, the detailed design is carried out, and the system development is completed. The system is proved to be convenient and powerful, which makes bank training more intelligent, efficient and economical.
【学位授予单位】：浙江理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TP311.52

【参考文献】