chapter eight
8 Inferring user activity from Android accelerometer data
This chapter covers
- Visualizing positional data from your phone in 3 dimensions along with time
- Performing exploratory data analysis and identifying patterns in Android phone users
- Automatically grouping Android phone users using their positional data using clustering
- Visualizing K-means clustering
Nowadays we are pretty much inseparable from a small thin usually black device that connects us to each other and to the world: our mobile phones. These devices are computing marvels: miniaturized chips with powerful microprocessors that are much more powerful than desktop computing from a decade ago. Add to that capacious connections to WiFi networks that allow broad connectivity to the world and Bluetooth, which allows narrow and close by secure connection to edge devices. Soon, WiFi 5G and Bluetooth 6 will increase these connections to geographically disparate networks and to terabytes of data and to millions of interconnected devices making up the Internet of Things (IoT).