Application Of Deep Learning In Virtual Experiment Guidance And Modification

With the rise of the Internet strategy to the national strategic level,the integration of "Internet +" in the field of education industry has opened a new chapter in information education.The "Internet + Education" has also brought new opportunities and challenges to the virtual experiment teaching and practice.However,due to the experimental courses being conducted online,problems encountered by students in participating in the experiment can not be promptly guided,and teachers are required to timely review the experimental results,resulting in heavy workloads,These have restricted the application and promotion of the Internet virtual experiment teaching system.This paper first investigates the status quo of domestic and foreign research on virtual experiment teaching and automatic modification system,introduces the automatic modification and intelligent guidance system of the original virtual simulation experiment teaching center for electronic information,points out the defects still exist in the virtual simulation experiment teaching center,Proposed the research goal and the research content of this article.Then it introduces the related concepts and theoretical foundation of this paper in detail,and introduces the principle,function and common calculation model of neural network and deep learning.Finally,the intelligent scoring system based on virtual experiment scene analysis is designed and implemented.The system has got rid of the shackles of teachers' making standard answers and experiment rules in the past when the system of reformation and guidance needs teachers.It has fully tapped a large number of effective experimental cases accumulated in the teaching experiment in the past,obtained useful information from them,established a deep learning intelligent scoring model,Discipline virtual experiment has the characteristics of versatility.The dissertation completes the design and development of virtual experiment automatic scoring system based on deep neural network.System scoring results objective and reliable,functional and intelligent.Verified and tested under real data,received good results.Practice has proved that the intelligent scoring system implemented in this paper is feasible.It is hoped that through the research of this paper,it can help the virtual experiment learning and teaching,improve the practical ability of students to meet the teaching needs of teachers at different levels.