I'm relatively new to data science, and I decided to stretch a bit and try a project with time series data. I downloaded the Human Activity Recognition from Continuous Ambient Sensor Data Data Set, and read the first file into a Pandas dataframe.
That dataset contains relative time data:
lastSensorEventHours lastSensorEventSeconds lastSensorDayOfWeek 0 10.0 38464.0 4.0 1 10.0 38465.0 4.0 2 10.0 38578.0 4.0 3 10.0 38582.0 4.0 4 10.0 38582.0 4.0
I've done some work with absolute time series--for example, analyzing stock data, using the time/date data as the index. With this, I'm not even sure where to start. I'm thinking to group into last sensor event length of time, but I don't know how that could be valuable.
Can anyone recommend some methods for how to tackle this data as a time series problem?