The first in a three part series on machine learning experimentation at Sift Science. In this post, we discuss minimizing bias in offline experiments.
