Joining the Machine Learning Master's Programs

When applying, applicants choose between being considered for admission to the three-semester (16 month) Master of Science in Machine Learning program, the four-semester (20 month) Master of Science in Machine Learning - Advanced Study program, or both programs. To enter the Advanced Study program, students must have indicated their interest in that program in their application and have been selected for it by the admissions committee. A student's offer letter will specify if they were admitted to the Advanced Study program.

Students who are given the choice between the three- and four-semester programs need to choose their desired program when they accept their offer of admission. These are treated by the university as fully separate degrees, and students can't switch between them after matriculation.

On this page, "MSML" will apply to both versions of the program, with "three-semester" referring to the coursework-based program and "four-semester" referring to the Advanced Study program.

Degree Title

Graduates from the three-semester program will receive a diploma that says "Master of Science in Machine Learning."

Graduates from the four-semester program will receive a diploma that says "Master of Science in Machine Learning - Advanced Study."

Who Should Choose Each Program?

The three-semester program is ideal for students pursuing a career in industry. Students can focus on courses taught by our world-class faculty, preparing them to go on to successful careers in areas such as big tech, quantitative finance, and other high-demand fields. At just 16 months, the time and cost commitment are shorter while providing an excellent education. Students who do an internship for their summer practicum can also often secure a full-time return offer for January, since employers know they will be graduating after just one more semester.

The four-semester program is ideal for students considering pursuing a Ph.D. The extra semester's worth of research allows a deeper investigation into whether full-time research would be something they would enjoy, as well as honing their skills so they're prepared to hit the ground running when they begin their dissertation research as doctoral students. By spreading the coursework over four semesters instead of three, students can also complete more research before applying to Ph.D. programs in their third semester, making for stronger Ph.D. applications. The May graduation date also aligns well with a standard August start date for a doctoral program.

Coursework and Practicum

The coursework and practicum components of the two MSML degrees is almost completely the same. Information can be found on our curriculum webpage.

For context about CMU's course system, graduate-level courses are 12 units, and a full-time courseload is 36 units each semester.

All students must complete all six core courses. All students must complete a summer practicum, which is either research or an internship related to machine learning.

All students must also complete three electives, which can be either courses from our electives list or machine learning research.

The only difference is that for students in the three-semester degree, if research is used to fulfill an elective requirement, students register for 12 units of 10-620 Independent Study, which is graded with a letter grade. For students in the four-semester degree, if research is used to fulfill an elective requirement, students register for 12 units of 10-630 Master's Thesis Research, which is graded Pass/No Pass and is expected to be related to the student's master's thesis.

In both cases, only 12 units worth of research can be used towards the elective requirements in each semester.

Research

In the three-semester program, research is completely optional. Students who want to maximize their research in that program can fulfill all three electives and the practicum requirement with research. Students who want to do no research can instead take nine classes and do a summer internship. Since research is not mandatory, this also means that the Machine Learning Department does not guarantee that every student will find a research advisor.

Students in the three-semester program are limited to 12 units of 10-620 Independent Study each semester. 12 units counts as one elective.

In the four-semester program, research is strictly mandatory. Students must complete at least 36 units worth of research, along with having the option to fulfill electives and/or the practicum with research. Students must write and present a master's thesis. Since research is a mandatory component of the program, the Machine Learning Department will assist students who have trouble finding a thesis advisor if necessary. Students are still expected to make a good-faith effort to find a research advisor when they arrive on campus, to ensure a mutually satisfying advising match.

Students in the four-semester program can complete 0, 12, or 24 units of 10-630 Master's Thesis Research each semester. They don't need to indicate in advance whether those units are being used to fulfill the elective or the master's thesis requirements. Students must begin research no later than the second semester, must continue it through their third semester, and must present and submit their thesis in their fourth semester.

Residency

Since MSML courses are so rigorous, MSML students are limited to taking 36 units in their first semester. Students may be allowed to overload with advisor approval in later semesters. As a general rule, however, most students find the MSML coursework to be so challenging that they ultimately decide not to overload, so we encourage students to assume they will only be able to fit 36 units into each semester.

The three-semester program is officially 16 months, consisting of Fall/Spring/Summer/Fall. The most common graduation date is December. However, students in the three-semester program have the option to complete the program part-time with advisor approval. Most courses are offered during the day and all courses require students be physically present on campus, so taking courses part-time while working is very challenging, although it is possible with a flexible employer. International students should be aware that student visas require full-time enrollment and that student visas are only valid for the length of the program (in this case, 16 months), so all students in the three-semester program with student visas must graduate within three semesters.

The four-semester program is officially 20 months, consisting of Fall/Spring/Summer/Fall/Spring. All students will graduate in May. Students in the four-semester program must be enrolled full-time in all four semesters and the summer. It is not possible to be enrolled part-time, and it is not possible to graduate early.

Admission Into CMU Doctoral Programs

Neither program guarantees admission into any doctoral program at Carnegie Mellon. Students who complete the M.S. program are welcome to apply to the Machine Learning Ph.D. program, but will not receive preferential treatment.

Finances

The tuition and fees for both programs are the same. Students are expected to pay their own tuition in both programs.

Funding from faculty is rare and cannot be sought before arriving on campus. Teaching Assistantships aren't possible in the first semester, and they pay an hourly rate that wouldn't cover the cost of tuition. Students should expect to pay their own tuition without departmental or advisor support.

Visas for Foreign Students

The length of program for students in the three-semester program is 16 months.

The length of program for students in the four-semester program is 20 months.

F-1 student visas do not allow for part-time enrollment, and they do not allow students to extend their studies beyond the standard length of their program.

Based on current laws, students in both programs are expected to be able to use Curricular Practical Training (CPT) for a summer internship or research in the U.S. between their second and third semesters. They're also expected to be able to use Optional Practical Training (OPT) for post-graduation work in the U.S., if desired. Information can be found on the Office of International Education (OIE) webpage.

Post-Graduation Outcomes

Since this is the first year the four-semester program is being offered, there aren't any alumni yet for us to have program-specific outcomes.

Historically, approximately one-quarter of MSML students tend to be interested in pursuing a Ph.D. after graduation, while three-quarters aim for industry. We expect that the four-semester program will be popular among students aiming for a Ph.D.

CMU’s Career & Professional Development Center (CPDC) has compiled first-destination statistics of our alumni. (A small caveat: If someone earns a master’s in December and works in industry or as a research assistant for half a year before entering a Ph.D. program in August, they’re listed as “employed” instead of “continuing their education,” since that is their “first destination.”) They publish similar statistics for summer internships, for students doing an internship instead of research.

Questions?

If you have any questions or would like help deciding, feel free to reach out to Machine Learning Master's Programs Manager Dorothy Holland-Minkley (pronouns She/Her or They/Them; calling them by their first name is standard in CMU's collaborative culture).

Dorothy has managed the MSML for a decade and is here to help. All questions are welcome!