In general, and unfortunately, applying to graduate school can be a dizzying and confusing process. Each year, I receive emails from ~100 prospective students who wish to apply to our programs and work with me. I’m flattered and excited that you’re interested in Data Science, Machine Learning, NLP! I read and respond to every email, but I’ve decided to write this webpage, in order to address many of the common questions and concerns — and to help detail the general process for universities in the USA. Hopefully this answers most, if not all, of your questions. If you have further questions, please contact Daniel Weinstock.
Again, I am thrilled that you’re interested in applying to Harvard’s Institute for Applied Computational Science (IACS), which centers around two Master’s programs:
Data Science
Computational Science and Engineering
PhD programs
A PhD is a research degree and typically lasts 5-7 years. It completely centers around your research work with your adviser. Your adviser needs to have funds to help support you; PhD students receive a stipend (i.e., salary). It’s not the case that every professor has the resources to admit and work with new students every year. Typically, most professors have approximately 2-6 PhD students at any given time, meaning, on average, they "admit" roughly 1 student every year or so. The “admit” is in quotes because while a particular professor can strongly influence the PhD admission process, the overall admissions are determined by a committee(s) that helps ensure the department is balanced and in a healthy state.
Master’s programs
Master’s programs last 1-2 years and, depending on the university, slightly vary in their nature. That is, some are more research-focused (e.g., UCLA, where I went), but most are heavily class-focused. Thus, it’s not expected for every Master’s student to work with a professor on independent research. With that said, professors always encourage and get excited when students want to get involved in research — myself included! Master’s students have to pay tuition to attend, unlike PhD students. In some rare situations, Master’s students are able to get their employer or other organization to front the bill. Nonetheless, because you are not supported by and tethered to a particular professor, admissions are selected independently from any given professor’s funding options, availability, and interest in working with you! This is a good luxury to have. So, admissions are determined based on your application, and once you are enrolled, if you can find a professor to work with for extracurricular research, that is great!
IACS
Specifically, IACS' application deadline is in December each year, and admissions are determined by a few committees. Admitted students who choose to enroll start school the following Fall semester (i.e., 9 months after the application deadline). We strive to maintain fair, inclusive, and unbiased admissions. Since I serve on one of the admission committees, it is inappropriate to ask me or anyone else “Can you please admit me to your program”. I encourage you to apply, and all applications receive a thorough and fair consideration.
In December 2018, we received:
1,021 Data Science applications and admitted 86 (55 enrolled)
421 CSE applications and admitted 43 (26 enrolled)
In December 2019, we received:
1,161 Data Science applications and admitted 76 (34 enrolled)
569 CSE applications and admitted 44 (24 enrolled).
In December 2020, we received:
1,759 Data Science applications and admitted 102 (66 enrolled)
884 CSE applications and admitted 60 (40 enrolled).
We appreciate the immense interest!
If you're admitted and choose to enroll, I look forward to meeting with you, and I'm always excited to work with our students on various Machine Learning and Natural Language Processing (NLP) projects! I currently have availability to work with Harvard students and to guide Master’s students on their thesis (a thesis is not required to graduate; approximately 10% of our students select the Thesis option, whereas most choose the Capstone course option).
My Research
My work is within NLP, specifically discourse. If you’re interested in working together, please visit my homepage for a list of my current research projects, which includes work undergrads, Master's students, PhD students, post-docs, and professors. In general, my main lines of research include:
coreference resolution
entity linking
annotation
commonsense reasoning
American Sign Language
I’m always excited to hear your ideas for new projects!
Each semester, I give a talk on general post-bachelors plans, which includes many tips for Master's and Phd programs, career options, etc. Slides here.
Good luck in the application process!