The code is released for the paper Understanding Student Procrastination vis Mixture Models, Educational Data Mining 2018.
Jihyun Park (jihyunp@ics.uci.edu
)
July 2018
Written in Python2.7
.
Python packages numpy
, scipy
, random
, and matplotlib
are needed to run the code.
test_data.csv
: Sample data (simulated data) to fit the Poisson mixture model. Each row in the file is considered as a daily activity count vector for a student. 400 rows exist in this sample data.
demo.ipynb
: A quick tutorial of using the code.
pmm.py
: Code for fitting Poisson mixture model given a count matrix. The file has two classes--PoissonMixture
for the model andPoisMixResult
for storing and plotting the result.utils.py
: Has helper functions for calculating log probabilities.