Prediction Of The Career Development Direction Of College Students Based On Convolutional Neural Networks

In recent years,with the deepening of the state's emphasis on higher education,colleges and universities across the country have expanded their enrollment year by year,and the number of graduates has continued to grow.The increasing employment groups have brought huge employment pressure to the society.Therefore,to improve the professional ability level of college students,to provide students with personalized career development direction planning,to help students establish a correct concept of career selection,is a feasible way to solve the structural contradiction of College Students' employment.With the deepening of digital campus construction and the development of Internet technology,a large amount of education data has been accumulated but has not been reasonably used.Therefore,this paper takes the massive student data accumulated in the digital campus as the research object,based on the research method of convolutional neural network,excavates the students' personalized information hidden behind it,forecasts its future career development direction,and provides scientific and technical support for the education and teaching work in Colleges and universitiesBased on the above background,this paper takes the data of 853 students from 2010 to 2018 as the research object to carry out the research work.Through literature analysis,a large number of literature materials are consulted to determine that students' academic performance,English level,political civilization activities,scientific and technological innovation activities,cultural and art activities,and social voluntary service are the influencing dimensions of College Students' career development direction,and on this basis,database sources are selected.The collected data are cleaned,integrated and transformed to form a standard data set for subsequent experiments.Firstly,the convolution neural network technology is used as the starting point for in-depth study,and the sequential model algorithm based on keras framework is selected to build the prediction model of College Students' career development direction with one-layer network and six kinds of eigenvalue tags.The prediction accuracy of the model is evaluated,which proves that the prediction effect of the model has the value of further in-depth study,and then the prediction model is further optimized Turn.Then add three-layer network and SGD optimizer to the model,and use soft Max as regression function to test the prediction effect of the optimization model.Finally,based on the optimization model,design and develop the prediction software for the career development direction of college students.Based on the real data of the graduates of the school,analyze the actual application of the prediction model.The average accuracy of the prediction model is 81.48%,the standard deviation is 4.34%,the ACC value of the model is stable around 0.835,and the loss value is stable around 0.3,which proves that the prediction model has a good prediction effect.Then,based on the prediction model,this paper designs and develops the prediction software of College Students' career development direction,which improves the efficiency and applicability of the prediction model.It provides a set of application mode that can be used for reference for the combination of convolution neural network related knowledge and career development direction prediction of college students.