but do you need to access all 1.3 trillion rows to do predictive analytics? why not just use a very small sample of the data?
but do you need to access all 1.3 trillion rows to do predictive analytics? why not just use a very small sample of the data?