MOOC Dropout Prediction Using a Hybrid Algorithm Based on Decision Tree and Extreme Learning Machine

Algorithm 1

Feature Extraction Processing.

Inputs:

: Learning behavior records of a course

: Enrollment number of learners

: Number of behavior features

: Duration of the course

Outputs:

: Feature matrix with size of

: Label matrix with size of

1: is the set of learning behavior records for each course. It is grouped by

the behavior types. Let be the record set of

learning behaviors, where .

2: Divide the duration of this course into weeks.

3: For each learning behavior record in

4: If this record occurred in week generated by learner

5:

6: For each learner, the learning behavior feature

is obtained.

7: Form the feature matrix , where is

types of behavior features of the learner.

8: Form the label matrix , where .