Jump to content

Recommended Posts

Posted (edited)

Hi everyone,

 

I would appreciate if anyone, who has been in the same situation, may give me advices. I recently got off the UMich waitlist and received the admission into their PhD Stats programme. Now, I am fortunate to be able to choose among top programmes in Statistics.

A brief background about me, I am a CS undergraduate. However, my concentration is CS Foundation & Theory so I am actually more comfortable with writing proofs and reading research papers than coding tedious programming projects. My research interest is Statistical Machine Learning. Career wise, I want to work in the academia to eventually become a professor and I am not very interested to join the industry.

I understand UMich Ann Arbor Phd is a very good programme, but I also have option to enroll MS in UChicago or Harvard then try again next year for the top 3 PhD programmes. The latter approach is certainly risky. Nevertheless, will it be worth the cost ? Is there any advantage that PhD graduates from top 3 programmes have over the ones from top 10 such as UMich ?

Edited by machinescholar
Spelling
Posted

I'm not in the same position but I'm interested in ml so maybe my advice can help a little. I'd say definitely take the Michigan offer (unless you're still waiting on other programs). They have a ton of people working in ml (Ambuj Tewari, Shuheng Zhou, Ji Zhou, Elizaveta Levina, Yuekai Sun, Long Nguyen, Susan Murphy, Clay Scott, probably missing some) plus they seem to have a good connection to the cs department where there're even more people doing ml research. It's hard to imagine you wouldn't find somebody who's research you liked there. It's also top 10 so it's not like you're really missing out much on prestige compared to the others in the top 5 (except maybe like Berkeley or Stanford). Since you already know you want to do ml research and you think you want to go into academia, look at the placements of the past students of these professors doing ml research, I'm sure there are a good amount who went into academia.

Going to Chicago or Harvard for an MS doesn't guarantee that you'll have better results next time around (after all better here means getting into one of <10 schools since Michigan is already really highly ranked). You could be spending 10s of thousands of dollars to attend a school with a slightly better reputation (or attend a school of the same reputation as Michigan or worse).

I also wouldn't fixate so much on "top 3". While Stanford and Berkeley are obviously some of the top stats programs for ml, Chicago isn't as strong in ml (Lafferty is the only really big name w/ a few other newer professors, although you do have access to TTIC), and other programs like Washington, CMU, Wisconsin, and honestly probably Michigan (just to name some in the top 10) are probably better than Chicago for ml.

Unrelated to your question, when you say you got off the waitlist, do you mean you were notified that you were on a waitlist and then taken off, or that you just received a second round admit?

Posted (edited)
35 minutes ago, marmle said:

I'm not in the same position but I'm interested in ml so maybe my advice can help a little. I'd say definitely take the Michigan offer (unless you're still waiting on other programs). They have a ton of people working in ml (Ambuj Tewari, Shuheng Zhou, Ji Zhou, Elizaveta Levina, Yuekai Sun, Long Nguyen, Susan Murphy, Clay Scott, probably missing some) plus they seem to have a good connection to the cs department where there're even more people doing ml research. It's hard to imagine you wouldn't find somebody who's research you liked there. It's also top 10 so it's not like you're really missing out much on prestige compared to the others in the top 5 (except maybe like Berkeley or Stanford). Since you already know you want to do ml research and you think you want to go into academia, look at the placements of the past students of these professors doing ml research, I'm sure there are a good amount who went into academia.

Going to Chicago or Harvard for an MS doesn't guarantee that you'll have better results next time around (after all better here means getting into one of <10 schools since Michigan is already really highly ranked). You could be spending 10s of thousands of dollars to attend a school with a slightly better reputation (or attend a school of the same reputation as Michigan or worse).

I also wouldn't fixate so much on "top 3". While Stanford and Berkeley are obviously some of the top stats programs for ml, Chicago isn't as strong in ml (Lafferty is the only really big name w/ a few other newer professors, although you do have access to TTIC), and other programs like Washington, CMU, Wisconsin, and honestly probably Michigan (just to name some in the top 10) are probably better than Chicago for ml.

Unrelated to your question, when you say you got off the waitlist, do you mean you were notified that you were on a waitlist and then taken off, or that you just received a second round admit?

Thanks for the advice marmle !

There was no official email but I knew I was on the waitlist because I emailed my PI last week when the results rolled out.  

Edited by machinescholar
Posted

Barely related but michigan has a big data summer institute summer program. I'd think that supports the idea they are a good school for machine learning. Personally turning down a top 10 offer (if you like the school) to spend thousands for a top 5 is way risky

  • 1 month later...
Posted
On 2/16/2017 at 0:50 PM, marmle said:

Chicago isn't as strong in ml (Lafferty) is the only really big name w/ a few other newer professors

Lafferty is apparently leaving Chicago.

Posted (edited)

Yea, he's definitely leaving Chicago. My friend was considering their Stats program, and he told me that when he visited, a professor told him that Lafferty is going to Yale next year. He was really disappointed because he wanted to work with Lafferty. 

Edited by flabbergasted
Posted (edited)
9 hours ago, flabbergasted said:

Yea, he's definitely leaving Chicago. My friend was considering their Stats program, and he told me that when he visited, a professor told him that Lafferty is going to Yale next year. He was really disappointed because he wanted to work with Lafferty. 

Indeed, that is where he is going. Chicago started to build momentum on the ML/computational front with him. I wonder how the department is going to respond with this set back?

Edited by arima

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...

Important Information

This website uses cookies to ensure you get the best experience on our website. See our Privacy Policy and Terms of Use