Application of fuzzy neural network algorithm in the analysis of learning behavior of university users

Abstract

The rapid development of Internet technology and education informatization has accelerated people’s learning and changed their way of thinking and cognition. Colleges and universities have set up their cloud classroom learning platforms, and the mining and analysis of the platform’s learning data is, therefore, particularly important. In this study, the SOFM-based FCM algorithm is used to perform fuzzy clustering of user subjects with different learning behaviors, and then the FDMA algorithm is used to mine the association rules of users’ learning behaviors, and the fuzzy neural network algorithm is also used to predict the learning performance to achieve the fuzzy neural network algorithm based on the fuzzy neural network algorithm to analyze the learning behaviors of users in colleges and universities. On this basis, we are developing and implementing a system to monitor the learning behavior of college users to explore its practical value. Based on different user behaviors, college student users are divided into three categories: high motivation to study (29.8%), medium motivation to study (46.2%), and low motivation to study (24.0%). The confidence level of the mined fuzzy association rules takes the range of [0.84, 0.95], which has a high confidence level. The academic performance prediction model had an average relative error of 0.0278 and 0.0281, with better model fitting and higher accuracy of prediction. The success rate of user access to the system is high, and the system has been well-tested. This study provides a reference for the application of fuzzy neural network algorithms in the analysis of user learning behavior in universities.

Keywords

• sofm
• fcm algorithm
• fdma algorithm
• fuzzy neural network
• learning behavior monitoring system
• 68m01