Subscribe to DSC Newsletter

Which masters or PhD program should I choose for Data Science or Artificial Intelligence?

 

Which masters or PhD course should I choose for Data Science or Artificial Intelligence?

 

This is one of the most common questions I am asked

 

Here is my view – and the usual caveats apply – i.e. the viewpoint is personal and is not associated with any company or institution I am associated with

 

Before addressing this question, you should also read two articles by me

 

Advice to a fresh graduate for getting a job in artificial intellig...

 

And

 

Landing a 150k usd datascientist job

 

The common theme here is: I take a market driven view to data science.

 

For instance – the data scientist is more valuable the more directly they are involved in creating intellectual property for their company. 

 

When you first start finding masters and PhD programs for data science or AI, you realise that most universities do not have a dedicated AI department.

 

And that’s probably a good thing in my view.

 

The impact of AI / Data Science is too sweeping to be centralised.

 

But that also means that its not easy to choose a masters / PhD program

 

Here are my suggestions

 

  1. Look at the top 25 Universities and choose a quantitative masters / phd program even if it is not directly data science related. The methodology for choosing top 25 universities is strongly biased to an emphasis on research
  2. Look at Universities performing strongly on Neurips. For example, see this analysis of neurips stats 2019 . Duke University and University of Southern California are high in these rankings
  3. Choose a masters or a PhD in areas which deal with large amounts of data(astronomy, climate, nuclear physics etc) – some of the best data scientists I know have a background in domains like Astronomy – ex Dr Kirk Borne
  4. Focus on AI/ML domains in niche areas – ex in the UK, Royal Holloway is well known for its cyber security program
  5. There are many areas of AI which are in the research domain – and centers of excellence are developing around areas in academia – for example -   Interpretable machine learning

 

So, why choose a ‘non AI’ subject to get into AI ? (ex nuclear physics or astronomy)

 

The common theme in the above approaches is .. you need to demonstrate the ability to solve large, complex, quantitative problems typically involving large amounts of data. This idea acts like an Occam’s razor or a proxy.

 

Another way to look at it is: to take the Elon Musk approach – what is the hardest problem you have solved  and your approach? So the PhD / masters should ideally end up working on hard problems and also learning the scientific method (approach)

 

When viewed in this way, your options are a lot more boarder

 

   

Views: 2289

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Videos

  • Add Videos
  • View All

© 2020   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service