1. literature study --- Online course grade predictive model based on behavior analysis - accounting for any wide Shang Fuhua

• Study

  Learning data Chongqing University of Arts "College Computer Basics" course of "play class network" platform
• Research motivation
  
  For a more scientific analysis of the behavior of online learning online courses and accurate forecasting results, this paper presents the results online courses forecasting model based on behavior analysis.
• Literature review

           The study is using data mining techniques, students collect data and conduct online learning neural network to implement a model to predict the results of an online course.

• Study Design

  • 1. The model building include: data acquisition, data analysis, data processing of the entire process performance predicted.
  • 2. Online learning behavior analysis: including behavioral components analysis, behavior and interaction behavior analysis operation three aspects.
  • Experiment 3: 1) Data Analysis and Processing

                  2) algorithm: Before the experiment using the results to predict the completion of pre-training and post-forward neural network.

                  3) Predicting Performance: Select 1000 freshman behavioral data, to 5: 2: 3 ratio, respectively, as a subset of the training, validation and testing of a subset of a subset of behavioral data.
          
• Using data sets
  
  Select "play class network" platform of Chongqing University of Arts "College Computer Foundation" back-end database as the experimental data

   

  Analysis conclusion

  Learning experience

  Learning behavior data acquisition can be obtained through data mining techniques, the latter part of the algorithm needs to remind ourselves of the neural network.