In a sample data file, I have Stream columns showing decisions based on the data from a single data stream. The Overall column shows the output of a black box non-ml algorithm based on past/current stream data. The Truth column shows when someone was actually present in that particular area of the home.
I am going to analyse the data and see if I find anything interesting and be able to generate my own detection output that approaches the Truth and if I can make any conjectures about the black-box algorithm used to get the Truth column.