Sorry for this question but i really don’t understand the first step to combine 2 train set
do i have to group by the combined dataset by id? merge by column? by row?
What kind of 2 train sets?
Try to analyze each country without any combining. I’v got my current result just eleminating features from house hold data. There are a lot of unnecessary information, which in its pure form only worsens the result.
1 Like
There are 6 filles for train in total. 3 countries and then for each country one file at household level and another at individual level.
Assignment is to work at household level. But you can build new features by doing aggregations from individual data.