Skip to main content

A postdoc in a US institution with letters from UsUS faculty can "reset" much of the potential disadvantage from doing a Ph.DPhD from a non-Us university.

Having said that, in addition to the "familiarity" issue that @aeismail brings up, there is also the issue of logistics. The effort involved in bringing someone who is outside of US over for an interview if they are not in the US is high, so the potential expected payoff bar gets a little higher. In your case, since you'd be in the US, this would no longer be an issue.

But getting back to the basic question: There are a number of foreign universities that are "well known" in the US, and applicants from those schools will not be perceived as weaker in any way. Also, for specific subject areas this can be even more specific (i.e a weak unversityuniversity might have a strong specialization in topic X, and so students working in X will be highly rated).

As for how many postdocsPostDocs you need, since you work in machine learning (i.e computer science at large) you should read the "best practices for postdocs" document that the Computing Research Association just put out.

A postdoc in a US institution with letters from Us faculty can "reset" much of the potential disadvantage from doing a Ph.D from a non-Us university.

Having said that, in addition to the "familiarity" issue that @aeismail brings up, there is also the issue of logistics. The effort involved in bringing someone over for an interview if they are not in the US is high, so the potential expected payoff bar gets a little higher. In your case, since you'd be in the US, this would no longer be an issue.

But getting back to the basic question: There are a number of foreign universities that are "well known" in the US, and applicants from those schools will not be perceived as weaker in any way. Also, for specific subject areas this can be even more specific (i.e a weak unversity might have a strong specialization in topic X, and so students working in X will be highly rated).

As for how many postdocs you need, since you work in machine learning (i.e computer science at large) you should read the "best practices for postdocs" document that the Computing Research Association just put out.

A postdoc in a US institution with letters from US faculty can "reset" much of the potential disadvantage from doing a PhD from a non-Us university.

Having said that, in addition to the "familiarity" issue that @aeismail brings up, there is also the issue of logistics. The effort involved in bringing someone who is outside of US over for an interview is high, so the potential expected payoff bar gets a little higher. In your case, since you'd be in the US, this would no longer be an issue.

But getting back to the basic question: There are a number of foreign universities that are "well known" in the US, and applicants from those schools will not be perceived as weaker in any way. Also, for specific subject areas this can be even more specific (i.e a weak university might have a strong specialization in topic X, and so students working in X will be highly rated).

As for how many PostDocs you need, since you work in machine learning (i.e computer science at large) you should read the "best practices for postdocs" document that the Computing Research Association just put out.

Source Link
Suresh
  • 51.1k
  • 5
  • 124
  • 247

A postdoc in a US institution with letters from Us faculty can "reset" much of the potential disadvantage from doing a Ph.D from a non-Us university.

Having said that, in addition to the "familiarity" issue that @aeismail brings up, there is also the issue of logistics. The effort involved in bringing someone over for an interview if they are not in the US is high, so the potential expected payoff bar gets a little higher. In your case, since you'd be in the US, this would no longer be an issue.

But getting back to the basic question: There are a number of foreign universities that are "well known" in the US, and applicants from those schools will not be perceived as weaker in any way. Also, for specific subject areas this can be even more specific (i.e a weak unversity might have a strong specialization in topic X, and so students working in X will be highly rated).

As for how many postdocs you need, since you work in machine learning (i.e computer science at large) you should read the "best practices for postdocs" document that the Computing Research Association just put out.