Design and Implementation of Student Behavior Big Data Analysis and Prediction System Based on Deep Learning

With the continuous acceleration of digital campus construction, students have accumulated a lot of data and information in their daily study and life, which provides an important basis for student management. In this regard, based on the current situation and shortcomings in the practical application of colleges and universities, this paper will put forward a set of construction scheme of college students' behavior analysis and prediction system, aiming at constructing the corresponding student behavior analysis model and providing decision-making reference for multi-dimensional, multi-perspective and diversified educational programs. The whole system is designed with B/S architecture, the front end is an interactive interface, and the back end is based on Spark, Hadoop and Tensorflow to complete the construction of the server, which reduces the difficulty of use and improves the work efficiency. Practice has proved that the system supports the distributed parallel processing framework to optimize the convolutional neural network model, so as to improve the training and running speed of various analysis models, and can support users to use the student behavior analysis model to complete the analysis and processing of massive data, so as to make the school teaching and student management more accurate and scientific, and then make a useful attempt to promote the high-quality development of colleges and universities.