berkeley take out the trash day

[sdub]

hey guys, we're all going through this together. different schools are looking for different things and many applicants accepted to one top school are denied by another. sometimes it does seem like a crapshoot and that a little bit of luck and good timing is often involved. although we're probably all anxious about acceptances there is no need or reason to put down/single out other applicants. just put in your hard work and let the schools decide.

[bernard]

I felt like I have to say something here. I am sharing some of my experiences here. I apologize for going off-topic.

I think even though one has first-author publication, as an international student, you must bring something extra, like very strong LORs coupled with very strong grades. If I had done my undergrad in USA and I was US citizen, then I probably can get into any top university with just a first-author publication, and normal LORs. (I was told that in 90% time, undergrad do not have first author publication when they apply). It's not fair standards are higher for international students, because schools must admit very few of them due to enrollment limit.

It definitely helps if your supervisor has connections with profs at top schools. By connection, I mean whether or not your prof is known at the top schools. In two of schools I got accepted and at Berkeley where I interviewed, I remember my supervisor told me those schools contacted him before or after acceptance to let him know about my offer. One school even called him to check on me before accepting me. So having someone at top school who know your prof can be helpful.

But then, working for a good professor is not enough. You have to stick with him for a long time to get a strong letter from him. I believe no one can pull a first-author publication without working for about 2 years in his lab. In my case, I sticked with mine for 4 years, and ended with a first-author transaction paper, a very very strong letter, according to my interviewer. (I didn't get accepted at Berkeley not because I'm not qualified, but because the prof did not know I was international student until near the end of our talk.)

Those are my thoughts. I come from a top 10 school in Canada, and my school was a bit unheard of to some people. I have to do way more to show that I am as capable as someone from great school like UM, for example. I also did poorly on verbal and did not do any subject test. I hope my post helps future applicants.

There of course will be people who are jealous. Just ignore them.

[belowthree]

Well I've written two NSF grant proposals (okay technically one that I wrote again to resubmit) and I got rejected most places, so why don't you all quiet down and play nice.

(I'm pretty sure I'm not helping much with this flamefest...)

michigantrumpet: I think everyone knows you must have done some pretty good work if you ended up getting in all the places you did. Clearly there's more there than just a name on a paper and I'm sure your recommendation letters spelled it out quite nicely for the admissions committees. Anyone who thinks those letters were important just because of the name and not because of what they said about you probably isn't worth arguing with.

As for the rest of you... life's tough, deal with it? It sucks to get rejected, but get used to it, you're going to get rejected a lot for conferences, grants and all kinds of things. It's just how it is. (And if you aren't getting rejected, your work is probably not as innovative as it should be...)

[alice]

Okay, so one last comment.

I just do not want anyone saying "I guarantee the publications are not at conferences anywhere near this calibre" or "if you aren't getting rejected, your work is probably not as innovative as it should be...".

I know the weaknesses of my application, and I was not arguing someone's research is less valuable.

But you are not qualified to judge my research. In fact, you do not even know me, so please refrain from such comments.

I think the discussion here confirms numerous studies, that people behave much "less nicely" when anonymous.

[defmacro]

I felt like I have to say something here. I am sharing some of my experiences here. I apologize for going off-topic.

I think even though one has first-author publication, as an international student, you must bring something extra, like very strong LORs coupled with very strong grades. If I had done my undergrad in USA and I was US citizen, then I probably can get into any top university with just a first-author publication, and normal LORs. (I was told that in 90% time, undergrad do not have first author publication when they apply).

I have a first author publication in a top security conference, strong letters and I got rejected from Berkeley and Stanford. It isn't clear cut, ever.

[leogk]

Take it easy, guys....

These days the number of extremely well-qualified applicants far exceeds the number of slots in all sub-fields of CS. No question about it. I would say the decision looks like a random sampling. Really tiny difference may be a critical point of the decision. And, no one can say that the one who gets admission are always more qualified than one who gets rejection.

Life is not always fair to everybody .....

[finest_engineering]

Whoa. hold on folks.

michigantrumpet: Dude, I am not in any way trying to implicate that you don't deserve your admits. I don't know you, I don't know about your work or your field. That was not my point at all. I was responding to someone else asking about my stats because I was not rejected from Berkeley along with some others. I provided them and simply made the point that my stats being good didn't get me in everywhere last year, and probably won't this year. I don't at all think that they accept non-deserving people to these schools. I am merely saying that when their is 10,000 apps for 100 spots (or something like that) that means 100 rejects for every admit. It is impossible (given how busy these faculty are) for there to be much difference between the last 50 in and the last 50 out. My point is simply that there a lot of great candidates that don't get in and there is no simple explanation why, its just the way the cookie crumbles. My point about the third-author was to point out that their criteria cannot be reduced to 'stats', not that someone without a first-author paper shouldn't be getting in. Congrats on your tremendous admissions success, once you get into CalTech you'll have flopped a royal flush.

bernard: it isn't easy for US students either, even with first-author papers.

[michigantrumpet]

Dude, I am not in any way trying to implicate that you don't deserve your admits.

I appreciate that. I agree that admissions can be unpredictable. It's just a little insulting for my admissions to be reduced to complete randomness. Let's just drop it.

As far as first-author publications, I guess things are different depending on the area. I stand corrected. In systems/computer architecture, many of the projects can 12-18 months, tens of thousands of lines of code, and week-long simulation runs. I don't think too many undergrads can end up first-author on a project like that, but I could be wrong. It's the same reason most of the potential advisors I've spoken to say to plan on 6+ years for PhD since that's typical for my area at those schools (without a prior MS).

[finest_engineering]

I am in systems and I have written a first-author paper. I found the research problem, thought about it, came up with an algorithm of my own that solves the problem. Then I realized the algorithm was awesome because it applies to an entire class of problems. I formulated this class, came up with a rigorous proof of the optimality of the solutions the algorithm computes, and simulated it. The professor I worked with was an 'advisor' in the truest sense of the word- all my ideas, all my work, all my writing. That is what first author means. There is a huge difference between this and jumping in as a third or fourth author with a bunch of other grad students and a professor. I should know because I have been listed as a fourth author on a paper before. I did a lot of 'work' to get that fourth author nod, but the grad students and professor were calling the shots and the project was their baby. Michigan I am confident you will have this experience soon, it is a very rewarding one to go from a general problem to your own solution, all on your own.

I disagree about your 18 month window. This is probably true for the sort of projects you have worked on as an undergrad, as it was for me too. This is because undergrads are never brought into a project unless there is a lot of grunt work and unoriginal coding to be done. In reality, and IN PARTICULAR in architecture and systems, a problem can be conceived, investigated, and solved in a couple months easily. This is generally the case in mathy type stuff. It usually does always take 18 months to fabricate a chip or something, those ttype of people always have much fewer publications. In systems when you can just dream up an algorithm, throw together some math, and do a simulation, not nearly as much time is needed. Your sub field might be different, but your statement was rather general.

I am bugged that I actually know how to research independently, and a school like Stanford didn't even look at my paper (I can tell from download stats on the URL I gave them). THis is on top of crazy stats and what I consider to be a good essay. Well wish for a miracle for me!

[alice]

My paper was in game theory, and it took me a summer to do the work. I found (or came up with, depending on your philosophy) a class of games, proved several properties about them, introduced a stability measure and proved more things about it. The conference reviewers considered the result "important" and a starting point for future research.

I do agree that systems projects can take longer, because I worked in compilers for a year, and it can take forever to get the damn thing working (even when you know exactly what it's supposed to do). I did the entire programming and solved problems that appeared on the way, but it was my supervisor's initiative.

When we were almost done, it was running out of memory on medium size benchmarks. We added a garbage collection mechanism for our framework, that would not interfere with the rest of the compiler.

There will be a paper and a patent on that work (sometime, industry research moves slower), but it was too late for me in terms of grad school admissions. Also, sadly, my supervisor there did not have ties with those schools, so maybe that's why I sounded bitter. I didn't want to imply anyone can get good letters "just because".

[belowthree]

I just do not want anyone saying ... "if you aren't getting rejected, your work is probably not as innovative as it should be...".

Calm down, I wasn't judging your research or anyone's on here. I was just pointing out that rejection is something we all face at some point. I wasn't saying anyone's work is not innovative, but rather that any innovative scientist will eventually face rejection. If you haven't yet it doesn't mean you're not innovative, it means you're early on in your research career. Good for you if you haven't gotten a paper rejected yet, that's really quite good!

Nothing in my post was supposed to insult anyone are claim that anyone's research wasn't good. I was in fact, trying to say that any claims of such are totally ridiculous and pointless. My intention with that statement specifically is I was trying to point out that we should just all calm down and accept that sometimes people get rejected and sometimes it's fair and othertimes it isn't and worrying about it is a waste of everyone's time. Unless you honestly have a suggestion on how the admissions process could be closer to perfect that scales... in which case, suggest away, I'm sure many schools would actually be legitimately interested, most schools know that their admissions processes aren't that great.

It usually does always take 18 months to fabricate a chip or something

I generally agree with most of what you wrote, but I did want to quickly jump in and disabuse you of the notion that any serious chip fabrication project could be undertaken and completed from conception to publication in 18 months. That would be insanely quick and far below the norm for that kind of project. Some things really do take a long time and have research cycles on the order of years, especially if you end up need to snag a grant (and donated fab time, you'll need both) before you can even start, in which case you're looking at probably a year from concept to funding (if things go quickly) and then potentially years more for design, implementation, fabrication and publication.

[PCP]

Hi Alice,

I'm in game theory too. What sorts of things are you interested in? Where are you thinking of going?

You said you're at Waterloo, right? What do you mean your professors don't have any connections? I applied there too, and in terms of game theory at least, I know Kate Larson got her PhD at Carnegie Mellon w/ Prof. Sandholm.

[alice]

Hello PCP,

I worked with Kate. One of my references was from industry. Kate is indeed known for game theory/AI at CMU, probably less so at Berkeley or Stanford where I got rejected - or so I understand. I don't want to disclose more about my references, though.

Are you from Waterloo???

I worked on coalitional games. I only applied to the first 4 (+ Waterloo), and I regret that decision.

I should really go back to doing homework!

[PCP]

Oh, too bad. Waterloo isn't a bad place to stay though.

I'm not at Waterloo, but I basically applied everywhere that had a game theory group. I'm more in the theory game-theory community. Its strange, it seems to me that AI researchers study game theory, and theorists study it, and they don't talk to each other or cite each other at all.

[alice]

Yes, that's very true ( They don't even go to each other's conferences. At some US schools, except CMU, I applied for theory as the first choice, because I didn't see anything close to game theory done in the AI groups. Waterloo does game theory in both the CS and the C&O departments (different flavours)

[PCP]

Well I know Stanford has A.I. game theorists: Yoav Shoham publishes in places like AAAI and AAMAS, and Daphne Koller has done some game theory as well. Berkeley not so much though. And CMU has pretty good game theory in both theory and AI.

[finest_engineering]

belowthree: I was responding to Michigan who said a conference paper in 'systems or architecture' takes 12-18 months to get from conception to written. This is not true (at least in the general sense in which it was said) as I have generated papers in systems in shorter time frames. My statement about a chip, and most actual 'wet' science in general taking 18 months was meant to be taken as AT LEAST 18 months - so we are completely agreeing.

Hey games people! I am doing game theory now too. It is fun stuff. You should be proud of what you accomplished in a summer Alice! Are you still working in games (or planning to in grad school)?

I recently got into games and found out how to use a game theoretic approach to a class of problems (my prof. suggested the game theoretic approach to a problem, but I worked out the details and did all the subsequent work). I found the nash eq. and also have a decision process I can prove gets to the Nash from any i.c. I also coded up a simulation. In other words, enough for a conference paper and it took me only a few months. Unfortunately this work was after I submitted my app. I e-mailed the admissions people to add some notes to my file about my recent results, but I think they already somehow decided they didn't care about my application at that point. I feel so crossed that Stanford didn't even look at a single document (my c.v. or pubs) that I gave them the url for. For all the money I gave them I feel like I deserve for someone to actually download my freaking c.v. or LOOK at a first-author paper of mine.

[octav1an]

I feel so crossed that Stanford didn't even look at a single document (my c.v. or pubs) that I gave them the url for. For all the money I gave them I feel like I deserve for someone to actually download my freaking c.v. or LOOK at a first-author paper of mine.

lol, i did the same thing. Neither Stanford nor Berkeley looked at my documents or visited my home page

[alice]

finest_engineering: I guess you must be doing networks stuff... many game theory problems are inspired from systems.

Will likely do game theory in grad school - I like the game-like setting =). Simulations can help tremendously (even in a theory problem... the main insight came from there (in my case))

And I would say don't worry. Your profile is great and you will get recognized for your work, eventually by the very schools that now closed the door =)

And by the way it was great to meet you guys. Although seeing results posted on this forum can help one go insane, I felt better to talk about it.

[PCP]

I know some of us have contacted professors at various schools. Have any of you contacted grad students? It seems like they might have a better impression of what its like to attend the place. Anyone? What did you learn?

[finest_engineering]

Alice: nah.. I am EE, not CS. I know there is a lot of game theory in networks. I am more interested in the theory per se. I am not really doing EE anymore either though. The best description of the direction I have been going the past year is basically applied math.

Octavian: I had great GRE's, 3.94 undergrad, 4.0 grad, a lot of research experience, work experience, a thoughtful essay, TA'd a grad course in my field, and all that didn't merit them looking at my pubs or CV. Doesn't the $100+ I paid them entitle me to having my applications materials actually looked at? What happens if I ask for my money back on the basis of this issue? Perhaps I should wait until I actually get rejected =)

[jhrgejdobre]

finest_engineering: are you appl. for cs phd or ee phd? is your master's is ee or cs?

Which ee depts. do game thy research? I don't know any!!!!!!! did you work on game thy as a ug?

maybe they looked at your paper through their library system?

[finest_engineering]

They couldn't have because it hasn't been printed yet, and they certainly couldn't have found my CV there.

I applied for EE, not CS. However, you may have noticed that at many schools the EE and CS departments are a single unit. The work I do is not really EE or CS, but rather control systems so make of it what you will.

As far as EE departments doing game theory.. I know of professors in EE deparments at all of the following schools that do game theory: MIT, Stanford, Georgia Tech, CalTech, there are others but I don't want to look them all up for you.

Keep in mind 'game theory' isn't like 'architecture' (two words you'll see thrown around these forums). One is a tool and the other is an application area.