I am a recent MSc graduate of a program in Machine Learning and Data Science with a BSc in Economics, and currently looking for Ph.D. opportunities in machine learning. Given my BSc background and the fact that I was unlucky enough to enroll in the MSc during the lockdown, I didn't produce any publications in Machine Learning.
Many Ph.D. programs that I am interested in applying, not necessarily at top universities, require applicants to already have publications on "top" avenues. I think it is because of the competition around the world for doctoral studies, but I still cannot get my head around why in a program when you are supposed to learn how to do research you are expected to be a top published researcher already just to be considered.
Is it common that the PhD programs in Machine Learning require applicants to have publications at top venues? Is it mandatory to have publications prior to applying?