UCLA vs Purdue vs Toronto vs Oxford

[Submh]

I want to do a PhD in CS. Which one do you recommend and why? 

Personally speaking, I feel Oxford has better reputation. Its ranking is 3rd-4th in CS for several years. 

[Feanor]

Depends, of course, on your particular subfield.

[Submh]

15 hours ago, Feanor said:

Depends, of course, on your particular subfield.

It is AI and more specifically Machine Learning/Robotics

[knp]

Are these schools not all competitive enough that you should apply to all of them, rather than trying to pick your favorite first?

[Submh]

On August 19, 2016 at 6:28 PM, knp said:

Are these schools not all competitive enough that you should apply to all of them, rather than trying to pick your favorite first?

They are all good. But assume that I have an admission letter from all these schools. Which one do I have to choose and why?

[knp]

7 hours ago, Submh said:

But assume that I have an admission letter from all these schools.

Right, but that is a bad assumption. Within my little speciality, I had a best-in-the-nation kind of research resume: I still only got into three of seven PhD programs. Apply to all of them, try your luck, and then concentrate on finding your favorite after you've been admitted. Or, if you don't want to wait that long, try looking at the work coming out of each of the four: which one do you think produces the most compelling research in machine learning? Whether or not a program is "best" in information processing doesn't matter if the next one is the best in the facets of machine learning you want to pursue.

Moreover, being able to answer this preference question well is itself a criterion for admission. The two PhD programs by which I was rejected were the two schools at which I did the worst job of explaining why that program would be the best fit for my research. Your statement of purpose is a persuasive document: you're trying to convince the professors reading it both that your research is amazing and that that school is the best place for you to do it at the next level. I suggest that you learn a lot more about the professors with whom you'd be working at each institution, because if you can't convince yourself of the specific strengths of a place, you won't convince them to admit you.