A project I made for a Quantified Self class. We were asked to create something that kept track of our lives in some way, so I decied to use my smartphone as a window into my daily habits. Not by monitoring what apps I use or where I go, but simply by using the accelerometer and proximity sensor on the phone.
I used the GMM clusterer from the scikit-learn Python package to cluster the sensor data into various orientations: On a desk, in a pocket, in a hand, sitting, standing, etc.
Here are the slides for the presentation I gave