Mechanical engineer, with 3 yrs industry experience as a Business Analyst. What should I do to increase my chances of getting into a good ML grad school?

[Tanul]

Background-

The Good: - The 3 yrs BA experience (working in the analytics field has actually inspired me to go one level deeper and transition into ML)

The Bad: GRE: 319 (V:155, Q:164, AWA:4.0) TOEFL: 107 CGPA: 3.20

The Ugly: - No publications - Undergrad degree in Mechanical engineering

 

I have completed Andrew Ng's ML course on Coursera and I have the below mentioned tasks in mind before I start my applications for Fall '17

1. Kaggle - Pick any of the ongoing/completed competitions and showcase the methodology/performance comparison between models on any blog/github

2. Replicate results from research papers - I have found some really cool research papers and I really want to replicate their results with some modifications. But unsure if it would be worth the effort/time and if it would even help my case

 

Also, in parallel to these things I am planning to complete some more online ML courses to try and get a more comprehensive understanding.

 

Does this seem like a good idea OR is there something else that I should be doing? Do I even have a chance? Should I retake GRE? Should I give up looking out for grad admissions and go hunt for jobs directly?

 

I know this is a bit late wrt to the fall applications this year. But at this point I'm getting all out paranoid.

 

Appreciate the help!

[compscian]

Hi @Tanul I'm an ME who transitioned into ML, and can probably share some insights. 

1. When you say ML program, which department are you talking about? If you want to join a CS/EE or Stats department, you will get degrees in that discipline. Hence they will naturally expect familiarity and pre-requisities (eg. if you join CS/EE/Stats you might have to take operating systems / communications / measure theory). Do you have a background in these areas; and are you interested in doing these courses for sake of degree?
2. In absence of above, your best bets are "data-science" type programs at places like NYU. Alternatively, you can look into operations research or information management programs and take a good number of electives in Stats and CS.
3. Online courses are not useful by themselves for a number of reasons -- see this post by a professor http://qr.ae/1JprKQ
4. Kaggle competitions are useful on your resume if you actually accomplish something impressive. Participation means nothing since anyone can do that. To get good results in Kaggle, you need a good understanding of ML, and it's not easy. On the other hand, if you get impressive results, it will certainly add a lot of value, but it will be hard.

My suggestions: Retake GRE, maintain a good online presence with your codes on github, apply to data science, business analytics, or info management type programs.

Honestly, there isn't much you can do to improve your profile at the stage. One long shot is to join some professor at IIT or IISc as a research assistant and work for a year. However, convincing them to take you is not easy -- but strong interest in the form of MOOCs, Kaggle, and github repos might help. You can then use the professor's recommendation letter and apply for Fall 2017. It's a circuitous path, but is likely your best option if you want to get into CS or Statistics departments. 

[Tanul]

Hi @compscian, Thank you for the reply! Its great to hear about living proof that such a transition is possible! 

The details and some follow ups -

1. Eventually I want to do PhD around NLP, and the best bet that I feel right now is to get into a MS in data science course and then collaborate with the ML Prof in that Uni to get into research and PhD later on. I wouldn't want to apply to a MS in CS/Stats/EE course because of both zero background coursework and next-to-impossible chances of getting in. (Not to mention I'm not very inclined towards doing their coursework as well).

2. I would apply to MS in Data science and also try the Operation Research courses as well (Didn't think of the OR option, thank you for suggesting this!). The Mech undergrad can actually help out in my OR applications

3. I'm not doing the online courses to increase the weight of my applications per se, but more on the lines of improving my own basic theoretical knowledge around the subject.

4. I know this is tough as hell but as long as it is worth it, I'll do it.

I don't think I have enough time left to be able to build a strong online portfolio and then try and join research projects with a Prof here. Also, should this be necessary if I apply to MS in Data Science/ Operations Research programs? 

Also, I would like to PM you and directly hear more about your ME to ML journey.

Let me know and thanks again!

[compscian]

I am OK with you PMing me. For the more general questions, I can take them here so that others benefit too. For more specific or personal questions, we can try PM on gradcafe.

1 hour ago, Tanul said:

Eventually I want to do PhD around NLP

I would apply to MS in Data science and also try the Operation Research courses as well (Didn't think of the OR option, thank you for suggesting this!). The Mech undergrad can actually help out in my OR applications

Some statistics departments are joint with IEOR ones -- they are particularly good choices (e.g. Georgia Tech, Princeton etc). However, note that OR or data science programs will not help you towards the ultimate goal of PhD in NLP, for which you have to be in CS, EE (esp. for speech, sequence prediction, or dialogue systems), or Linguistics (semantics, language understanding). You'll run into the same problem of department mismatch after your MS, and your chances might be even lower than now (bar for PhD >> bar for MS).

If you are particularly interested in NLP, and also want to develop a parallel skill set in NLP, I'd actually suggest a computational linguistics MS program. For example UWashington has an excellent program. Otherwise, try and apply for "data-science" programs. Operations Research won't look good on your resume since it's very far from NLP.

[Tanul]

Good point! OR will again be very different and far off from NLP. In that case, I'll just treat OR applications as the last resort.

Any thoughts around this? 

11 hours ago, Tanul said:

I don't think I have enough time left to be able to build a strong online portfolio and then try and join research projects with a Prof here. Also, should this be necessary if I apply to MS in Data Science/ Operations Research programs? 

How important would it be to get a research publication for getting into MS in Data Science programs?