When #8 is finished, we will have access to 73 households.
We will choose appliances that are shared within many households, then split the houses in two groups: one for train, other for validate.
We will compute the appliance threshold on the train set, then compute the intrinsic error of those thresholds on the validation set.
Note that we do not need to train any model for this study.
When #8 is finished, we will have access to 73 households.
We will choose appliances that are shared within many households, then split the houses in two groups: one for train, other for validate.
We will compute the appliance threshold on the train set, then compute the intrinsic error of those thresholds on the validation set.
Note that we do not need to train any model for this study.