Masters in computer science or Masters in applied statistics?

[trickslapper]

Crap. I couldn't get everything i wanted in my topic title.

Anywho, I'm wondering which masters degree would be better for a career in machine learning / smart data while still keeping other - traditional options available? I would love to do a double major and earn both graduate degrees but I'm pretty sure that's impossible with a full time job (if it's not impossible it will take too long for me).

I've read a lot about this and I have mixed feelings. I've seen some people say that if you have an undergraduate degree in computer science then a Masters in Statistics/Math/Physics leads to a powerful combination.

What would you guys suggest? If it helps i have my bachelor degrees in computer science and math and currently work as a software developer (so i plan on going to school part time for right now)

 

[Theorem.]

My significant others Father is a professor at the University of Alberta in Computing Science. He publishes a lot in Machine learning. Formally, he is trained in Applied Mathematics and I think he considers this a huge advantage. He takes on a lot of graduate students and I think he is often frustrated with the level of math students enter with. Many students coming from computing science graduate programs lack the proper mathematics background and he spends a lot of time teaching them mathematics or giving them mathematics to learn. That is just one perspective though.

 

[trickslapper]

My significant others Father is a professor at the University of Alberta in Computing Science. He publishes a lot in Machine learning. Formally, he is trained in Applied Mathematics and I think he considers this a huge advantage. He takes on a lot of graduate students and I think he is often frustrated with the level of math students enter with. Many students coming from computing science graduate programs lack the proper mathematics background and he spends a lot of time teaching them mathematics or giving them mathematics to learn. That is just one perspective though.

What level of mathematics do you need to be able to have such a career in machine learning?

 

[Theorem.]

That really depends what you are working on in machine learning. He works on the theoretical side so as you can imagine, he requires a lot of mathematics. In general I think one should have a good idea of such topics as Statistics, Probability, error analysis, Linear algebra in order to study in Linear Algebra.
Check out http://courses.washington.edu/css490/2012.Winter/lecture_slides/02_math_essentials.pdf !

 

[trickslapper]

That link is awesome. I think i'll go through the rest of the pdfs on that website. But back to my original question.. do you think an MS in CS or Stats is better for a career in machine learning? I would love to get both of them and sometimes i lean towards the Stats and then other times i lean towards the CS.

 

[Theorem.]

That link is awesome. I think i'll go through the rest of the pdfs on that website. But back to my original question.. do you think an MS in CS or Stats is better for a career in machine learning? I would love to get both of them and sometimes i lean towards the Stats and then other times i lean towards the CS.

Do you hope to work on the more theoretical side? Then maybe statistics is better. The people I know have a PHD in either statistics or probability (or combined). But It's not like it isn't possible with a Masters/PHD in comp sci. The professor I know takes on many masters and pHD students on topics related to machine learning through the CS department. A neat option (possibly very beneficial) is to get a masters in statistics/probability (you could even focus on something related to machine learning), and then go for a pHd in a computer science department !

 

[trickslapper]

I don't hope to work on the theoretical side. But i would like to know some theory so i sort of understand what I'm doing when i do hands on work (hopefully one day). I added a couple of classes to my coursera watchlist that relate to machine learning/data science so maybe those will help me make a better decision.

 

[Theorem.]

THose will, and you should ask the professors too for those classes if you get a chance. Knowing the theory can be a huge advantage, although my opinion is a little biased- coming from pure Math. Good luck : )