Research On Programing Ability Evaluation Method Of College Students Based On BP Neural Network

The college students who is the social reserve talents must master a certain degree of skills.Especially for computer science major students,the curriculum requires strong practicality,and it requires the students have strong practice ability.However,many college students have insufficient theoretical knowledge of computers and lack of actual combat capabilities.It is difficult for IT companies with high demand to recruit suitable software talents.It is fundamental to solving the software talent gap to studying how to bring up and improve college students' programming skills,and shortening the gap with the needs of the company's employeesAt present,there are few studies on the evaluation of software developers' programming ability in our country,and the evaluation of college students' programming ability is still lagging behind.There is a lack of scientific programming ability assessment model.Designing a comprehensive and accurate evaluation model of college student'sprogramming ability can promote the informatization of education and teaching management.In addition,it can feedback the level of student's programming ability in real time.On the one hand,it can also provide students with feedback on learning effects.On the one handand it can provide reference for employers to recruit talents as well.Therefore the studies have theoretical value and practical significance.This paper first introduces the research background and significance of the project,and briefly summarizes the research status of domestic and international capability evaluation.According to the characteristics of software developers,through the literature review method,expert rating method and other factors that affect the ability of programming is analyzed.Finally,the programming ability evaluation index system is established.Then,combine with the advantages of fuzzy analytic hierarchy process and BP neural network.The fuzzy AHP method is used to determine the initial weight of each evaluation index and the comprehensive evaluation value of programming ability.And then the training and test sample output of the BP neural network programming ability evaluation model is determined.As a result,by statistically analyzing college students' online programming behavior data and constantly adjusting network weights and thresholds.A dynamic assessment model of college students' programming ability based on BPneural network was established,and the prediction of college students' programming ability was predicted.Finally,the simulation experiment shows that the error between the predicted value and the expected value is within the acceptable range.The feasibility of the evaluation method was verified by simulation experiments.Finally,the programming ability evaluation system needs analysis,the main functional module design,the system implementation and other three aspects of the elaboration is accomplished.And designed the evaluation system in the behavior monitoring,behavior statistical analysis,programming ability assessment and other three major functional modules ? At last,the preliminary design and implementation of the evaluation system for university students' programming ability is completed.It provides a practical evaluation tool for the evaluation of programming ability of computer professional majors.