You will need following packages
- graphlab-create==1.5.2
- matplotlib
- scipy
- pandas
- numpy
Create a data folder in the dir where the notebooks are run and copy the csv file in this data folder.
Start Ipython
$ipython notebook --pylab inline
Notebooks
- exploratory-1.ipynb
this ipynb does some basic data analysis on the available dataset
- exploratory-2.ipynb
basic statistical analysis to segment different levels of user activity based on country, device.
- activity-classifier.ipynb
this notebook contains steps to build a boosted tree classifier on the acitivity data
- svm-activity-classifier.ipynb
construct a support vector machine classifier for activity based on engagement metrics data