nathan and audrey

Well, I have been pretty much incognito for the past week due to my first attempt at submitting (it was just an attempt to submit for a chance to publish) a paper to a conference. The punchline to the story: I failed to complete the paper in time. It is extremely disappointing to work so hard for something and see it come to no where, vaporize into thin air if you will. I have been working about 14-17 hour days for the past week. It is truly amazing how much we can achieve given the right motivation. I didn’t know I could live reasonably off of so little sleep, although, I will not be making it a habit to get 3 hours of sleep every night. Even though I kept working under such conditions next time I will be better prepared (maybe I can’t say this, one thing I should learn is it never goes how you think) for the amount of work.

It is important to note that I was almost completed the paper. I have now learned and am in the process of sorting out how to handle time management of huge deadlines for the future. Read on for a description of the events.

Summary of Events

I would like to provide a brief description of what happened so everyone can know what I’ve been up to for the last week.

My team and I worked really hard last semester on this research as a class project. When we completed the semester our T.A. and prof suggested that we continue on with the work because they thought it had the potential to get a publication.I was in awe at that suggestion, I couldn’t believe that something I did could be publishable. Now of course that is the goal of going to grad school, but hey, I’m still just a small guy trying to get his first publication.

So essentially what happened was that everyone went their way for break and left the project. When we got back I had emailed the T.A. for suggestions on conferences to aim for. He sent me an email with several upper level ones, and suggested we make an attempt. The only problem with this was that some of the most desirable conferences had due dates in February, and we only had a crude implementation completed, as well as no data or experimentation yet. So off we went frantically trying to collect data an analyze it for the best results.

One thing that happened during this initial phase of work was that we didn’t have an advisor to help guide us. I was basically pushing the research forward with no idea of what to focus on. With this being my first publication attempt I felt really hesitant to move forward. I was unsure of wether we were spending our time on the right stuff and extremely unsure of some assumptions we were making for our research to work. Thankfully about a week ago I finally got some time with my professor and we discussed the project. To be honest that came to nothing, primarily, because we didn’t get the paper written on time, but it was still a huge hurdle. The same day I also met with our T.A., and he provided me with loads of good questions and topics that we should cover. After this confirmation of a direction I became really motivated to get working on the project. I started reading the previous research papers and started to see how our work was really novel. I noticed how the assumptions we were making in our research were the same assumptions that previous works had made, and not only did we have the same assumptions but our research improved on some of them.

As I started to read more of the previous works I really got excited and convinced that we had the “good stuff.” That we were guaranteed a publication. Hmm… no publication without a submission though . Nevertheless we charged on with renewed vigor. I was reading and writing our paper while my teammates were writing code to collect the thousands of data observations we would need to be successful.

Time-wise in the story we have gotten to Sunday, and still hadn’t collected any data. The paper was moving along smoothly, but it would hit a wall very soon without results. I incorrectly allowed myself to take time to work on the paper when I should have focused on the data collection code. Well, it turned out that on Monday one of the students became to busy to work on the project and I took the load of writing the code. So I spent several days frantically trying and testing out tons of methods to collect this data. Eventually on Wednesday night I had code it up and running on several machines collecting the data. This was an amazingly Euphoric feeling! I had made it! We were going to succeed because by the next morning we would have the stuff! See the image below for my data collection sweetness. Neat huh?

Well, that feeling faded fast as Thursday afternoon rolled along and we received our first set of results: 4% success rate!!!! I wanted to cry. At this point I really started doubting whether or not we could get done on time, but I pressed onward. I performed some analysis and improved our results to 23%, which was more than enough to push for the submission. The last day (yesterday) was a tumultuous attempt to collect these results and perform additional tests while writing the paper. The code to perform our experiments took a long time to figure out. I didn’t really start writing the final portions of the paper until 5pm, the paper was due at 1:59 a.m..

Throughout the rest of the night I went back and forth asking myself if I should continue to try and get done. I knew very easily that I wouldn’t be submitting my best writing, and that was unnerving. I didn’t really think it was wise to submit a poorly written piece of work. It was extremely frustrating though to have put forth such huge efforts and come up with no submission. At 2 a.m. I finally put the typing away and went to bed.

Lessons Learned

Throughout the process I have learned a lot of new things. I will be parsing these as I go along over the next several weeks, but I thought it would be a great idea to capture some of them right now.

• Time management.
• Results FIRST before you start writing the paper.
• It always takes longer than you expect on a tight schedule. Thus, plan less and start earlier.
• Getting bad results suck.
• Keep pushing even when it looks bad, and importantly when it looks good.
• Knowing when the data you have is enough, and setting a stopping point that will provide enough time to make conclusions and write the paper.

Summing It Up

This week was insane. I learned so much about my abilities to focus and work hard and how failure really sucks. I’m going to provide forthcoming posts that will discuss some of the lessons learned from this past two weeks, as well as some philosophical debate on whether or not one should submit a work that is poor in quality. These are in general more broad grad school questions that deserver their own post. In fact I had intended to place them in this post, but as the length grew I have decided to take them out and give my readers a break!

Oh, and as an astute reader pointed out, not all is lost. I have two more upcoming conferences to submit to, and now I have the right amount of time to get the paper where I really want it.

If you have done research or dealt with deadlines like this, let me know how you handled these types of issues.

Okay so I just posted on the new move, now for the thing that has taken up most of my life for the past week and a half.

I wanted to write a little bit about my experiences in the first few weeks of grad school, and reflect on what I expected in coming here.

The first thing to say is that I had no idea how much activity grad school would involve. My question this past spring as a pre-grad student was always, “What will my daily schedule look like? Will it require 20 hours of research and 2 classes? … etc.” I have since realized that those are not the right questions to ask. If you are really wanting to get an idea of how grad school is from the perspective of being an undergraduate student you must realize that the grad school life is a tremendous paradigm shift. I tried to understand grad school from a perspective of how many hours I would work, and have realized that the grad school experience doesn’t fit into a certain number of hours a week, it completely fills every minute minute of your life. .

With that said it isn’t bad, and you do have time for a lot of other things, but the idea is that when you are in graduate school you are applying yourself to your dreams. You are getting to do what you want to do. Instead of being told do this so you can get a job where people tell you what to do, you are asked, “what interests you?”

Another thing that changes is that to be a grad student means you adhere to a much higher standard than  you did in undergrad. You can skip pre-requisites for courses because you are expected to either know the information before class or learn it on your own. The great and awesome aspect of this is that it is your choice. You are responsible. No one else. I know that the difference between high school and undergrad is a similar one, but the jump from undergrad to grad is an order of magnitude higher.

So to sum up this concept, in grad school you work hard, but you become so motivated about that work that it isn’t bad. It’s like if you want to be a professional video game player what are you going to do? Play tons of video games! You will not be very good to start and spend tons of time at it, but in the end its extremely rewarding (don’t know about that part yet, but the mere suggestion of such an outcome fascinates me). Of course there are some lines that must conform too, but overall it is so freeing.

As per the last paragraph I have found shifting into grad school extremely easy and soooooo rewarding. I have been reading more than I ever did in my undergraduate years. I skipped a class and went to the library and read tons of pages the week before school. Why? Because I didn’t know the information and wanted to. No one made me, or said I should. I wanted to do and did. Soooo sweet! I have found the experience so far to be extremely freeing. This freedom has helped me to be motivated. Its hard work but great. My schedule has been drastically changed… Right now I basically follow this schedule:

• 7:30 Am get up
• 8:30 leave for school on bike
• 9-6ish: go to classes, workout, research, and meetings at school
• 6-9ish: hang with audrey
• 9:30ish to 12: do more work

Its been crazy, but awesome!

To sum up it is a crazy ride, and one that isn’t for those who aren’t prepared or that don’t really want to do grad school. I hope this post serves two purposes: 1) to remind myself of the feelings I had before coming to grad school, and 2) to open the minds of potential grad students so that they can ask the right questions in determining whether or not to pursue grad school, and which school fits them best.

As usual thanks for reading!

-Nathan

Hey Everybody-

The following is a video I recorded to discuss a bit about my graduate school selection. There will be a more detailed analysis in a post to come about the princeton visit. Thanks for visiting.

–nathan

This is somewhat of a hard post to write about because in a large sense I don’t have direct experience in what graduate admissions offices are looking at with respect to your background. I do know that your prior research experience can be the difference between you getting accepted or rejected. Some of this section will come from the guide: http://www-2.cs.cmu.edu/~harchol/gradschooltalk.pdf written by Mor Harchol-Balter. She has a much better perspective on what is a positive and negative from the admissions office viewpoint.

To get into the top graduate schools you need the following:

1. Good GRE scores
2. Good GPA
3. Prior Research
4. Great letters of recommendation

The first two of these are not as important when competing at the top schools. The reason why is because most of your graduate life will be focused on one thing: research. Therefore, the most important factor at the top institutions is how good of a researcher you project to be. This projection comes primarily from your past research experience. So the key to looking good for the admissions office is to prove that you have great potential as a researcher. A 4.0 and 800s on the GRE are not good enough at this level.They may be required to hit the bar minimum, but those alone will not get you in.

The question is then how do I get research experience? There are several options which include: (the numbers and main sections here are as provided by the previously mentioned grad school talk, but the descriptions following the bold letters come directly from my own experience).

1. Undergraduate research with a professor. The relationship you develop with your undergraduate professors is extremely vital because not only will they give you research opportunities, but they will also be the people who write your letters of recommendation. I have had several great relationships with professors leading to great undergraduate research opportunities. I was able to perform departmental honors and get credit for said research. So the best option is to start discussing potential research opportunities with professor you know well. At the larger institutions the advisors carry a few undergraduate researchers, which is a good place to start. Find out what your school has to offer and go with it. The key to finding a research opportunity is to have a good relationship with the professor you want to work with. It’s important to note that I went to a somewhat smaller undergraduate department, which might have made it easier to find this connection with a professor. A better description of how to find a professor is here http://www-2.cs.cmu.edu/~harchol/gradschooltalk.pdf :

1. How to ask a professor to do research with him:
   i. Make a list of professors with whom you might want to work. Most professors have a web page with lots of papers and descriptions of their research. Take your time and read these.
   ii. Make a summary sheet about yourself. This is a one-page piece of paper which you will be giving out. It should have the following information about you:
   (a) Your photo. Professors can’t remember names, but they do remember faces.
   (b) Your name and contact info.
   (c) Brief description of your research interests.(d) Your availability – are you looking for a full-time summer position, a part-time fall position, both? How many hours can you devote to this project? Are you looking for class credit?
   (e) List of every class you took and your grade and the professor you took it with (professors like to talk to each other about you).
   (f) Relevant previous experience and skills.
   iii. Talk with your undergraduate advisor – at CMU this is Mark Stehlik – about whether anyone has an open position for undergraduate research. Mark maintains many such lists.
   iv. Pick a professor from your list. Go to his/her office hours or send email to schedule an appointment (most professors prefer office hours). Explain your situation and give your summary sheet. Be prepared for the following questions which he may ask you:
   (a) Do you have any ideas about what you might want to do research on?
   (b) Which of my projects are you most interested in working on?
   (c) Describe your math background.
   (d) Describe your programming background.
   With very high probability the professor will thank you for your interest, but tell you that he isn’t taking on any undergraduates this term. This does not mean that he hates you! Be brave! Try the next person on your list. You may have to try this 10 times. Hang in there. If the  professor does say yes, you need to be prepared with your own list of questions:
   (a) What is the goal of the project?
   (b) What exactly will be my responsibilities?
   (c) Whom will I work with? – e.g., will you work directly with the professor, with one of
   his graduate students, with another undergraduate?
   (d) What background material can I read to learn more about this project?
   (e) What textbooks can I study to learn more about this research area?

2. Summer Internships: This is not only a great opportunity for developing research skills, but it also gives students helps studetns to connect classroom learning with real world problem solving. As an engineer this seems to be a big gap to bridge. I don’t think I would be ready to do research had it not been for several internships that I have worked at over the last 4 years of undergraduate time. So start applying for research internships. I worked with a few DOE labs in my undergraduate career.
3. Get a Job: I have been a student intern and applied to grad school directly from undergrad. So I don’t know too much about this. If the job is research based it will be a lot more useful than any other type.
4. MS Student Project: you will work on some project as an MS student giving you research experience. This one as well, I have no grid for.
5. Work Alone on a problem: Find out an area of interests and start reading papers on that topic. If you find an interesting problem, propose a solution and start working on it. I have never done this and it is actually more closely related to how you get your thesis topic as a PhD student. So this might be hard, but if you are driven it will help out.

I hope that this guide is helpful to you. Getting research experience should be started as soon as possible. You should start at least 3 semesters before you graduate so that you can prove you have experience when you write apply.

Any corrections or addions would be welcomed in the comments.

–nathan

I recently applied to several graduate schools and thought it would be a great idea to help out those people getting ready to apply with a sort of how to and when to for graduate school applications. I made several mistakes and several good choices during the process. My mistakes centered mostly on the timing. This will be an overview of the graduate school admissions process and how to best navigate the intense amount of requirements associated with it. 

These steps assume that you, the reader, already knows that you want to pursue a PhD, and understand the implications of such a commitment.

In writing this guide I will pull from several sources including my own experience, but whenever I’m getting information directly from another source I will post a link to that site. My current list of sources include:

• Mor Harchol-Balter: http://www-2.cs.cmu.edu/~harchol/gradschooltalk.pdf
  • This guide is the one that I used most during my applications process. I’m going to use it and add my spin onto what it is talking about. It’s good to get the opinion and experience of someone who just went through the process. This guide is extremely thorough and worth the read.

I will be writing a series of posts to describe each of the following parts of the process:

1. Make yourself look good here
2. Deciding which schools to apply to
3. Creating a personal/research statement
4. Getting letter of recommendation writers
5. Fellowship applications
6. When you apply
7. Waiting for results

If I have forgotten an important step, comment and I will make sure to add a post for that. As I write these documents I will also add links from the list here to their respective pages. I assume as I go through the creation of the pages I will remember other things that must be discussed. I hope this is informative and useful.

–nathan