Design of Informatization College and University Teaching Management System Based on Improved Decision Tree Algorithm

Abstract

At present, the teaching management system used in colleges cannot classify and store the teaching material information well and also has some problems, such as inaccurate calculation results of resource information weight, long response time, and large data query error. Therefore, this study designs an information college teaching management system based on improved decision tree algorithm. The hardware structure of the system consists of information communication structure, information teaching resource sharing structure, processor, and crystal oscillator circuit, and the core module is the data output control module. This study designs the system software based on the improved decision tree algorithm, creates a decision tree recursively, uses the CART decision tree to calculate the weight of teaching resource information, and constructs the fitness objective function of teaching resource information according to the mean clustering algorithm, so as to accurately extract the teaching resource information and realize the efficient processing of college teaching resource. The experimental results show that the response time of the system in this study is only 8 ms, the maximum convergence value is only 40, there is only one wrong data in the data query, and the storage time is 40 s. This system has a short response time, fast convergent rate, and a low probability of data search error when more than one client access the database at the same time.

Impact of Covid-19 epidemic on online learning and educational resources in China

“…There are various research methods for predicting dropout behavior. According to traditional machine learning methods, the most commonly used methods for predicting dropout behavior include support vector machine [30], logical regression (LR) [31], decision tree (DT) [32], random forest (RF) [33], hidden Markov model [34], etc. Fei et al regard the prediction of dropout behavior as a classification problem about time series and use LSTM to encode the characteristics of learning behavior into continuous values and predict [35].…”