Hi,
If I understand correctly, each individual prediction on the test set need to be done on a 360 samples window of 1 device.
My intuition says that not all devices in a room will be affected from the presence of people in it, why the prediction need to be with only a single device data and with all the devices in the room?
Lion.
Posted by: lionb @ April 29, 2022, 9:16 p.m.* not with all the devices in the room
Posted by: lionb @ April 29, 2022, 9:17 p.m.Hi,
The train data set is more extensive than the test data set - more devices included in order to allow participants to train their model on different devices.
The test data set more closely resembles a real-world scenario where there are relatively few devices in the room.
MAFAT Challenge Team
Posted by: MAFAT_Challenge @ May 1, 2022, 10:12 a.m.