Application Advice - Biologist interested in AI

[atitus11]

Hi, I need some graduate school advice. I graduate in May with degrees in both Biology and biochemistry along with a minor in Math. I am really interested in going to graduate school in AI/Machince Learning/Natural Language Processing and I was wondering what I should do to make my application the strongest. I would love to go to a T1 grad program, and want to make my application as strong as possible. I am taking some time off right now to build the creds, and was wondering what I should be doing.

My Stats:

GPA: 3.7+

GRE: 790Q, 620V, 5.5W

Technical courses taken:

Intro Computer Science - Java

Calc I/II/III

Linear Algebra

Diff. Eq.

Statistics

Physics - Mechanics

Physics - E&M

Quantum and thermodyamics

I am currently in a research fellowship position at Stanford in a Neuro lab and will probably do about two years here. I also have 2+ years of neuro research experience from undergad.

Any advice would be great!

[emmm]

You seem well positioned to take some more CS classes and go to seminars/lectures and get to know some of the faculty. You should also be able to get some really good advice from the people you meet there. Good luck!

[gaiasabre11]

I agree with the reply above. IMO, you need to take more CS courses, unless you can prove that you have substantial programming experiences outside of school.

[atitus11]

Thank you for the suggestions.

I will hopefully be taking another course or two in the fall. Right now I am self teaching a bunch of the fundamentals and programming skills that will be needed.

Are self taught skills, as long as they are truly solid, acceptable for graduate programs?

[Pauli]

Are self taught skills, as long as they are truly solid, acceptable for graduate programs?

Very likely not, especially if you plan to take your quals and prelims in CS, where they'll be asking you an assortment of questions from core topics such as data structures. You may end up being tested on concepts such as sorting and trees, just two of many topics.

[tkulk]

in that case don't apply to CS schools. If you are interested in AI, you can take the neuroscience-AI route and enroll in the appropriate programs. For eg. Caltech's CNS program, MITs BCS program or NYUs CNS program. However, even for these programs, you will need exceptional quantitative and programming skills. However, formal CS courses may not be as much of a hard requirement as in the case of CS schools - research and self-taught skills (in form of projects) could be okay.

AS far as top schools are concerned, it's all about research potential and there are numerous ways of showing that.

Edited May 16, 2012 by tkulk

[Adamah]

PhD or MS? Etiher way, you would need at least Data Structures and Algorithms, and ideally Computer Architecture, to get into any CS grad school. You should find a way to take these classes before applying. tkulk made a good point that maybe you should look at specialty programs instead of general CS schools. CMU offers PhD's specifically in natural language processing or machine learning, for example. However, you'd have to show these schools that you know the subject well enough that you are confident you want a degree specifically in that area.

[atitus11]

Thank you tkulk and everyone else. Adamah - PhD :-D

I think that the Neuro-AI route sounds awesome. I have been considering that. I do have very strong quant skills, so I am not worried in that realm. As for the self taught stuff, there is a ton of coursework online, such as Stanford Engineering Everywhere (SEE) that provides the full course material including lectures, assignments, tests, and keys. By the time I apply I will have "taken" the full courses in Data Structures, Algorithms, Discrete Math, and a couple more, but I just will not have the actual credit.

AND, by the time I apply I will have been doing neuro research for 4+ years of mostly full time work. The department I get my PhD in does not really matter to me, as long as I am getting the training that I need to excel in my research field. I am pretty sure that is a ubiquitous byproduct of any top-tier research program.

[Adamah]

As much as I like online coursework, I don't think schools would count it. You might be able to get away with it if you take the CS subject GRE, but I'm not sure.

Also, you should see if you can take the courses for free, since you'll be doing full-time research at the university.

Edited May 16, 2012 by Adamah

[Elimba]

I suggest that you talk and get in touch with professors and graduates students active in the field of Bioinformatics. This is a highly demanded expertise on the job market today, specially in top research industrial laboratories. Machine learning (AI) and statistics are the keys to becoming an expert in bioinformatics. This is more about Computational Science than computer science as such. Therefore, you must get advice from your supervisor, choose carefuly your courses (from the palette of courses offered at Stanford) and give the right direction to your research. With the description of your background you gave us here, you are well on your way to mastering bioinformatics. You may also want to read the sound advice on the "PhD Candidate" page of the "Academic Joy" Web site. Good luck.

[atitus11]

Thank you Adamah and Elimba! I will look into both of your suggestions. This is all great advice!

[psycho_killer]

CMU also has a new "neural computation" PhD: http://www.cnbc.cmu.edu/pnc

[atitus11]

Awesome, and definitely on the top of my list.

[mathgeek282]

I don't really know so much about programming skills for what you want to do. I got into MITs EECS area 2 (AI) and I have only had an intro CS course. I had tons of math and stat courses though and a few research experiences that required programming. But all that is to say that I think AI is actually a field where math background is more important than a formal programming background...just my two cents.

Good luck!

[thenewcsstudent]

Hey guys, so I have a question now. I'm a genetics major and am minoring in CS. I want to apply to CS graduate school and thought maybe i should take the CS subject GRE as well. My stat's are:

Cum. GPA : 3.47

Minor GPA (CS) :4.00

Research experience: 2 years

I have three LOR's from three different professors working in bioinformatics field. One of the professor is a CS faculty memeber and the other two professors are associate professors in both biological science and computer science departments.

All three LOR's are really good. I'll be taking GRE during the summer. 

My question : Is there a really good chance I would be admitted for CS graduate program at a top tier graduate school? I wasn't sure since I'm really interested in getting a MS in CS and later working in bioinformatics and even maybe database management.