Detection of behavior change in people with depression

Afsaneh Doryab, Jun Ki Min, Jason Wiese, John Zimmerman, Jason Hong (Carnegie Mellon University)

https://drive.google.com/file/d/0BxF_mgbVSV6VS1QxTThMZTZJa3M/view?usp=sharing

The use of sensor data from a mobile phone to predict changes in major depression.

Data collected: noise amplitude (from microphone), location, WiFi SSIDs, light intensity (from ambient light sensor), movement (from accelerometer), device states (such as screen on/off, apps currently running, and the battery-charging state), daily call logs, and text messages from the phone. Features were extracted from these data streams to generate three main behavior parameters: social interaction, mobility, and phone usage. 

Depicting Emotional Range

Nick Barclay

I enjoy the way Nick was able to use simple shapes and two colors to elicit visceral emotions that are immediately relatable.