Competition: Predict dropout rates for MOOC participants
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.

MOOC Dropout Rates

This analysis was performed to compete in the 2015 KDD Cup. The objective was to predict whether a given learner would drop out of a MOOC in the next 10 days given a set of activities such as lectures viewed and assignements completed. There were datasets Objects, Dates, Training.


Variable Description
course id Parent course
module id Unique ID assigned to each module
category Category of module such as Chapter, problem, sequential, vertical, video, discussion
children Children modules
start Date the module became available to the leaner.


Variable Description
course id Unique course id number
from Course start date
to Course end date


Variable Description
enrollment id unique identifier for enrollement.
username Unique ID for each learner.
course id Course the learner enrolled in.


Variable Description
enrollment id Enrollement ID corresponding to the Enrollment table.
time Time the event occured
source Server or Browser
event The event the learner took. Examples include navigating to the object, completing a problem, or accessing a site
object Object ID corresponding to the the unique identifier in the Object table