DanielWarlock

It is not satire and it is a collaboration effort. Actually they may have eventually resolved to buy like 10 of these and put them in the prof's office for safekeeping because it is much cheaper if you buy a bunch at a time.

I must give my opinion on university computing cluster and cloud solution: (i) university computing cluster will require you to submit a batch file and you will potentially need to wait for a long time before your job is executed. And if they do provide interactive, resources allotted to you will be very limited; (ii) I definitely DO NOT recommend using cloud computing such as AWS even if you have like $200 free credits to burn through--it is nothing. The cost can be as high as $30/hr PER NODE for some luxury machines! They will also charge you on storage etc also. I used AWS last semester because the experience is better than university cluster. BIG MISTAKE! After 2 month my bill is $700+, and I haven't actually done anything substantial. So if you can convince your cohort to buy a commonly accessible GPU then do that instead. Sometimes department has spare funds so they buy stuff like massage chairs but such money in my opinion is better spent on computing equipment.

I always dream of one of those "monster machines" with lots of cores (e.g. 16 cores) whereas my own macbook has only 2 cores. Some "pro people" do have such machines and would even carry around GPU/FPGA boards. Some of my classmates recently have pooled resources together to buy Nvidia's new multi-instance GPU (A100) for $199000 a piece!

In this case, maybe you can ask your current supervisor to refer you directly? It is different from just writing a letter. Your prof basically sets you up with his colleagues so that they will look after your application. I know people who got in this way. Also maybe that one faculty you are interested in does not even take any students this year. If so, you are throwing money away by applying that school.

Harvard applied math is affiliated to the engineering school (SEAS) instead of art and sciences; like other engineering disciplines at harvard it is very small (only 8 profs). I believe it has nothing to do with the "real" math department.

Your course list is heavenly. You should get into most top programs barring possibly Stanford.

Harvard is committed to be open. No one knows anything other than that. We are all very worried.

They like to admit people with interesting background. I myself specialized in Monte Carlo algorithm and they thought I would be an interesting addition to the cohort. There are people who did astronomy, fashion design, environment protection, Italian language and all sorts of things that can apply data science techniques in novel ways. I thought you have good chance given your background is kind of unique although it is not guaranteed--you should make case of how you would apply stuff such as neural network or other ML things to your specialty (i.e. BA). Graduates from my cohort is highly successful which actually exceeds my expectation-- most of them went to Facebook, Google, Apple, Uber or top finance firms such as JP Morgan, Goldman Sachs, Mckinsey, Bridgewater etc., and several including myself remained at Harvard for PhD in CS or stats. I also got offers from Berkeley Duke etc. Now that I'm graduating, the outcome (both my own and my classmates) indeed exceeds my original expectation of what a master degree can achieve. Two years before I worked as an IT-type of guy in an unknown company maintaining spread sheet and emails etc. and never thought I would have progressed this far. This master degree really paid off (meanwhile it also almost bankrupts my family financially and put me in a hard spot for the future 5 years so you need to think twice if you are married). In the case where you have money, you should apply and do a master here.

Most data science programs are also professionally oriented. But there is lots of flexibility in the Harvard program, which I went. I successfully got into several good PhD programs this year. I would recommend the Harvard program if money and time is not a problem for you. But it is scary, really. You will be basically taking next 7-8 years of your life in order to do a PhD and you have already had a master--10 years in total! Are you married?

I think OP wants to move into "data science" which is hot now. But either way, I don't think Stanford will admit given there are only 8 admits GLOBALLY. I don't think your background is good for other programs either, at least not typical. UChicago - MACSS may be your best bet yet. If you did BA, then applying to an actual job (e.g. in a bank) should make more sense.

I don't understand it. Are you applying for a second master after the first one?

Your profile is not extremely strong but it is good considering you are applying to only UK masters. The MASt is a good preparation for PhD and is generally well known in academia. A majority of the applicants will be admitted if I recall correctly. But I'm not sure it is a good option for industry biostatistics.

Of course the OP has his eye on Stanford (especially with such a magnificent profile). And yes, a sure way to secure a spot there as a PhD is to attend their master program in statistics. Your chance is superb. Good luck.

ML is more like a buzz word. What is exactly the kind of thing you want to work with? Jianqin Fan is a top stats prof at Princeton. I think what he did now mostly classify as "ML". Tracy Ke, who is now a prof at harvard, was his student. I went to her lecture a while ago and it was quite packed because it was considered ML. But she worked on high dimensional lasso stuff which is very different from say CNN or other neural networks. To give you some perspective I took a "ML" class with Leslie Kaebling at MIT. She also specialized "ML". But what she did is RNN/MDP in robotics. This again is very different from what Jianqin Fan (as a statistician) is doing. It seems that your interest mostly lies in CS. Why aren't you apply to CS PhD in computer graphics? It is true that some students in stats do work in the kind of stuff you are interested in but not very prevalent.

Yes that sounds very good. In this case, you should continue your research with the mentor and apply to statistics programs with a theory focus: Stanford (Dembo) for instance. Who is your advisor and what kind of research he works on? In general, you should be clear about what exactly you want to do and what it is. In general, you should go to a pure math program to get a top-notch education in theoretical research even though your intended field is probability. You should not apply to statistics because you think it is less competitive than pure math programs.

I think you have an ok profile but you look like a pure math guy to me. Are you sure you want to do biostatistics? If I were you, I'd take more graduate level math courses and drop that applied PDE research for something more theoretical (such as abstract algebra which you seem to be really passionate about). Harvard, Princeton, MIT, Stanford, Caltech Cambridge, Brown all have top pure math programs. Why aren't you applying to pure math program at the schools you have mentioned? I really do think you have the potential to get into one of those and thrive there instead of settling for other things. You should also consider other top schools such as UCBerkeley, Columbia, UCLA for good measure.

I really admire your record and drive to learn. My understanding is that once you can "do real math" and get recognition from experts in the field (i.e. professors), admission will focus more on that (i.e. publication and recommendation letter) instead of your GPA, or math GRE grade. After all, you get into graduate schools to solve problems not to take courses or do standardized testing. Have you thought about contacting one or more pro mathematicians and let them give you some research-level problems to solve? I imagine that solving a few open problems (maybe just the easier ones) will be the easiest way (for you in particular) to secure a spot at the top graduate programs. Don't be too ambitious when you start. You will likely not be able to solve one of those "famous problems". Just try to derive more generalized results based on the published papers in your field. My personal experience is that you should form your own "philosophy" of a problem that is unique and un-obvious to others.This does not always happen but when it does, it is something that worth sharing with the community (through publication).

Stat110 is a class for first year college student so it is most likely very trivial for you. The instructor for 110 (Joe) has written up a book for graduate student in STAT 210 but has not published it. You can send an email to him and he will most likely share it with you. Also I would like to comment that the real analysis is used in proofs in inference class also (not just measure theory). A list given by Micheal Jordan in the reddit post although I personally think this is a out-dated. If you were an older person like Prof Jordan you will probably appreciate "history" of statistics more but I personally will read several of these but take time to read more modern stuff.

Just so you know, I was rejected at a bunch of "safety schools" such as Uni of Florida and U Minnesota but got into UC Berkeley Harvard Duke ect. And you are a way better applicant IMO. You must apply to schools that fit you in terms of level and interests rather than trying to be "safe". Talk to your prof at UCL to see if he can guarantee you a spot. If so, you can save a lot of money by not applying to other "safeties". Otherwise, your list is very good and I stand by my opinion that you will get into Stanford.

This is what I call a "Stanford profile". You should be able to get into Stanford. That being said, you should apply to some other schools for good measure.

GRE: 166 (Verbal), 168 (Math) GPA: 3.92 (UofT Engsci undergrad), 4.00 (Harvard CSE master) Quant GRE: NONE Research: Monte Carlo, Markov Chain theory Publication: NONE Courses: Took graduate probability and inference at master; not much advanced math classes Recommendation: strong letters, profs know me well. Admit: harvard, berkeley, duke Rejection: University of Florida, University of Minnesota Twin Cities, Columbia University in the City of New York, Wharton School of the University of Pennsylvania etc. Decision: I will remain at Harvard even though Berkeley seems to be a better place to go with more "big names". This choice is personal because I know Harvard better and it suits me well so far. Advice: I think it is more important for one to focus on thinking and problem solving than worrying too much about the institution they will be affiliated to. Contact: you can contact me with yufan_li@g.harvard.edu if you have any questions about application or would like to collaborate on a research project/problem. This forum has been helpful. Thanks!

"top quant hedge fund"--with only this, you can get into some of the top MFin programs such as Berkeley, Columbia and CMU. Of course, those are just cashcow programs but you also get the chance to talk to top profs at those schools, which will lead to some opportunities.

I think you will likely be rejected by all of these schools unfortunately. Your background is simply not very good. Also you are basically throwing your money into water by applying to UofT because they rarely admit international student without personal connection.

Nothing for me

Columbia is highly prestigious.