Model for CASAS ADL Interleaved Dataset

[asorici]

Why

We have tested the CONSERT Engine on the CASAS ADL Normal dataset where people perform activities in a specific order and do not deviate from it.
We wish to test the detection capability on the CASAS ADL Interleaved dataset, where activities can be performed either in sequence or interleaved (e.g. starting one activity, then starting a second, before coming back to finish the first).

What

The CASAS ADL Interleaved dataset monitors different activities than the ADL Normal one. The objectives are:

• understand the activity descriptions from the README, as well as the layout of the appartment from the same folder
• inspect the format of the sensor data (e.g. for this user) and see if the wsu_proj/wsu_parser.py file needs to be updated to parse the new sensor format. Consider creating a new parser if differences are significant
• model the missing basic ContextAssertions that capture direct sensor data

Notes

A branch named 57-casas-interleaved has been created from 35-update-casas-model and will be the working branch for these issues.