I was quite confused about why there were so few observations in the confusion matrix until I read that you took a subset of the data in the "future improvements" section. This should definitely be mentioned earlier when you outlined your procedure in the analysis section (also, if there was a reason for choosing exactly 1000, explain that as well!)
I was quite confused about why there were so few observations in the confusion matrix until I read that you took a subset of the data in the "future improvements" section. This should definitely be mentioned earlier when you outlined your procedure in the analysis section (also, if there was a reason for choosing exactly 1000, explain that as well!)