Intelligent Prediction Method of MOOC Learning Behavior Based on Kalman Filter
In order to reduce the dropout rate of MOOC and improve the quality of MOOC teaching,an intelligent prediction method for MOOC learning behavior based on Kalman filter is studied.IMX582 camera of Huawei Nova 5 pro is used to collect learning behavior data in the MOOC environment.The autoregressive integral sliding average model based on the collected learning behavior data is used to establish an observation model for MOOC learning behavior and convert it into a state space model.The calculation results of the state space model is used as the initial state vector for Kalman filter.By updating the ini-tial state vector through Kalman filter,intelligent prediction results of MOOC learning behavior is obtained.The fading factor and covariance matching criterion are introduced into the Kalman filter to compensate the uncertainty of the posterior covariance matrix in the Kalman filter and correct the intelligent prediction results of MOOC learning behavior output by the Kalman fil-ter.Experimental results show that this method has high accuracy in intelligent prediction.After applying this method,the dropout rate of MOOC can be effectively reduced,and the learning activity of learners can be improved.
Kalman filterMOOC learning behaviorintelligent predictionautoregressive integral slidingfading factor
万方数据
维普
