Disclaimer: The descriptions on this page are the personal opinions of the Machine Learning Master's Program Coordinator, not CMU or the City of Pittsburgh.

What is Pittsburgh like?

The answer depends on where you're from. The city itself is about 300,000 people (the 68th-largest city in the U.S.), while the metropolitan area is over 2 million (the 30th-largest metropolitan area in the U.S.).

If you're from a midsized city, Pittsburgh has everything you could want: arts (symphony, ballet, multiple museums including some that are free to students), sports (football, baseball, hockey, soccer), nature (rivers and parks with miles of biking and walking trails), restaurants, shopping, nightlife and more.

If you're from NYC, you'll likely be surprised that you can walk from CMU to residential neighborhoods where 10 p.m. is considered a late closing time for the local shops. If you're from LA, you might be surprised that walking or biking to campus is perfectly reasonable and many students, staff and faculty members do so all year.

For photos of the city, check out www.visitpittsburgh.com. For more detailed information, flip through the 2024 Pittsburgh Official Visitors Guide.

What is CMU like?

Unlike many top universities, CMU has a compact campus. A "long" walk between buildings is five to 10 minutes and usually doesn't involve crossing any streets. A student in the Machine Learning Department can attend most classes, seminars and research group meetings while never leaving the Gates-Hillman Center or the connected Newell-Simon Hall. Anyone who wants to go off campus for lunch can get to the restaurants on Craig Street with a seven-minute walk from Gates-Hillman.

While Pittsburgh has excellent sports teams, collegiate sports aren't a major focus of CMU's social life, nor does CMU have a party-school culture. Instead, large social events at CMU tend to be either geeky or artistic. One of the largest events each semester is Capture the Flag With Stuff, where hundreds of students play capture the flag with over a dozen varieties of magic items. The biggest celebration each year is Spring Carnival, where students enjoy carnival rides, eat typical fair foods, and check out booths (tiny buildings filled with art and games) built by fraternities and clubs that are designed from scratch each year around a set theme.

The Machine Learning Department itself is close-knit. We occupy the eighth floor of the Gates-Hillman Center, and students and faculty often have quick chats while getting (free) coffee or hot chocolate in the kitchenette. The M.S. in Machine Learning program shares a lounge/study space with the M.S. in Computer Science program directly across the street in the Tepper building, just a few minute's walk away. The weekly Machine Learning Tea is always well-attended by graduate students, faculty and staff relaxing and chatting over snacks, and students regularly enjoy getting together for on-campus board game nights and off-campus rock climbing.

For a view of campus itself, check out the CMU Visit page, the campus map, a virtual tour, or scroll through the CMU flickr.

What is housing like?

CMU does not have graduate student housing. Most graduate students rent apartments in either Oakland (the neighborhood CMU is in) or the adjacent neighborhoods of Shadyside and Squirrel Hill.

CMU's Graduate Student Association hosts a housing resources webpage, including a Graduate Student Housing Handbook with detailed information about housing costs, neighborhoods and how to find an apartment.

Can I fit research into the program?

About half of our students are engaged in research in any given semester. Along with research that's done purely for fun, three out of nine courses can be replaced by research, and the summer practicum can similarly be completed via research. This means that our M.S. students can complete the program  with the same schedule as our Ph.D. students, completing two courses plus research each semester and doing research over the summer. Check out our full curriculum.

How would I find a research adviser?

It's the student's responsibility to find a research adviser, but we do have a standard method. The easiest way to match with an adviser is to wait until you're here, which is how our Ph.D. students do it. While some programs assign advisers to applicants when they're admitted, we instead want to give new students the opportunity to meet all our faculty in person before deciding on an adviser. The main way we facilitate this is through the departmental orientation at the beginning of the semester, when most of our faculty give half-hour research talks. We encourage the faculty to make it explicit whenever possible whether they are seeking Ph.D. or M.S. advisees — or both. It isn't uncommon for an adviser to have a strict cap on the number of Ph.D. students they're able to advise (since that is a many-year commitment) while they can be more flexible on how many M.S. students they're able to advise.

After listening to the talks, you can then reach out to any faculty you'd be interested in working with to set up a brief meeting to discuss your skills and interests. If you and the faculty member agree that you're a good match, you can begin working together at any time.

When could I begin research?

The timing for an M.S. student to begin research depends both on the individual student and their adviser. A student with a strong and relevant research background may be able to jump into a project in their first semester. A student without that background or who's entering a new research area may instead spend the first semester sitting in on their new adviser's research group meetings and doing background readings instead of participating in a project right away. Waiting also lets the student complete three courses in their first semester, which is helpful for students who haven't had an opportunity to study machine learning before.

Are research assistantships available?

Research assistantships for M.S. students are fairly rare. Approximately 10% of our master's students earn some form of RAship in any given semester. The extent of this RAship depends on the grants that their adviser has available in any given semester. For example, some grants require that RAships only be given to Ph.D. students and so M.S. students aren't eligible for payment. Another grant might be able to provide hourly pay to anyone regardless of class level. A different grant might offer tuition remission in exchange for a semester of work. However, most M.S. students simply complete research for credit.

Do note that since you won't be able to form a research relationship until after you arrive, you'll unfortunately have to make your decision about whether to join our program without knowing if research funding would be available.

Are teaching assistantships available?

Teaching assistantships are much more common than research assistantships. After you've completed a course, you're welcome to apply to be a TA for it. TAs earn hourly pay and many master's students TA. Learn more about becoming a TA.

Can I talk to someone about the program?

The Machine Learning Master's Program Coordinator, Dorothy Holland-Minkley, is happy to talk with prospective students at any time. You can email her, and telephone and video chats can be arranged as well. (For anyone concerned about forms of address, "Dorothy" is most common, but "Ms. Holland-Minkley" is also fine if you're more comfortable with that.)

Several current machine learning master's students have volunteered to speak with prospective students as well. Email Dorothy to request to be put in touch with them. If there are any aspects you're specifically interested in (such as international student issues, research experiences or undergraduate major), you're welcome to mention that and you'll be matched with relevant students as much as possible.