Harvard SM in Data Science: Machine Learning & Tuition

Harvard University offers a Master of Science (SM) in Data Science through the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS). Students apply and enroll through the Harvard Kenneth C. Griffin Graduate School of Arts and Sciences.

The 48-credit, 12-course degree is full time and on campus, and takes three semesters. Machine learning is not required of every student by name. It enters the degree through a Data Science/Machine Learning requirement in which students choose three of four options, one of which is CS 1810 Machine Learning.

Last reviewed: October 2026

Quick Facts


Degree	Master of Science (SM) in Data Science
School	Harvard John A. Paulson School of Engineering and Applied Sciences, with admission through Harvard Griffin GSAS
Location	Cambridge and Boston (Allston), Massachusetts
Format	Campus, full time
Credits	48 (12 courses)
Culminating Experience	Capstone project, or thesis by approved proposal
Minimum GPA	No minimum published
GRE	Optional
Application Deadline	December 1, 2026 for fall 2027 entry
Tuition	$67,504 for the first year and $33,752 for the third term (2026–27)
Typical Length	Three semesters
STEM Designated	Yes

Official Master’s in Data Science page

Curriculum

Harvard does not offer a standalone master’s degree in machine learning or artificial intelligence through SEAS. The SM in Data Science is jointly led by the Computer Science and Statistics faculties and is the SEAS master’s degree with machine learning in its requirements.

Students complete 48 credits, normally 12 letter-graded 4-credit courses, consisting of:

3 courses in Data Science/Machine Learning, chosen from four options
1 computing course, chosen from four options
1 required ethics course
2 disciplinary electives, one in computer science and one in statistics
4 additional electives
1 research experience course, the capstone project

Each course must be passed with a C or higher, and students need a B average across the 12 courses. Transfer credit and Harvard Extension School courses are not accepted.

Data Science/Machine Learning

Students choose three of the following four:

AC 209a – Data Science 1: Introduction to Data Science
AC 209b – Data Science 2: Advanced Topics in Data Science
CS 1810 – Machine Learning
One statistics course from the list below

The statistics options are:

STAT 111 – Introduction to Statistical Inference
STAT 139 – Introduction to Linear Models
STAT 149 – Introduction to Generalized Linear Models
STAT 211 – Statistical Inference I
STAT 244 – Linear and Generalized Linear Models

Because students pick three of the four, it is possible to finish the degree without CS 1810. Students who want the machine learning course should plan for it in this block.

Computing

Students choose one course from:

AC 215 – Advanced Practical Data Science
AM 215 – Mathematical Modeling for Computational Science
CS 2050 – High Performance Computing for Science and Engineering
STAT 221 – Computational Tools for Statistical Learning

Ethics

AC 221 – Critical Thinking in Data Science

Electives

Students take one computer science elective and one statistics elective. They then take four additional data science electives, which can come from SEAS, other departments in the Faculty of Arts and Sciences, other Harvard schools, or MIT. Two of the four may be non-technical.

Up to six of the 12 courses may be at the 100 or 1000 level. One semester of AC 299r – Special Topics in Applied Computation, an independent study, may count as an elective.

SEAS does not publish a fixed list of approved electives on its degree requirements page, so topics such as deep learning, natural language processing, and computer vision depend on the electives a student selects.

Capstone or Thesis

The research requirement is normally met with AC 297r – Capstone Project.

Students may instead complete a master’s thesis. They must find a faculty advisor and submit a proposal by mid-April of the first year, and only proposals accepted by the Data Science faculty committee go forward. The thesis ends with a written document and an oral defense, and AC 302 replaces the capstone and two electives.

Cost and Tuition

Harvard is a private university and charges the same tuition to Massachusetts residents and nonresidents. The SEAS master’s programs in Data Science and Computational Science and Engineering have their own tuition schedule, which is higher than standard graduate school tuition.

For the 2026–27 academic year, Harvard Griffin GSAS lists:

Category	Tuition
First year (two terms)	$67,504
Second year (one term only)	$33,752
Per term, four courses	$33,752
Per term, three courses	$25,314
Per term, two courses	$16,876
Per term, one course	$8,438

Across three semesters, tuition comes to about $101,256 before fees. The per-course rates apply to part-time terms, which are allowed only after the first three full-time terms.

Students also pay a Student Health Fee of $972 per term and $2,477 per term for the Student Health Insurance Plan, plus a $35 Graduate Student Council fee each fall.

SEAS states that Harvard is not currently providing financial aid for its master’s programs. A small percentage of second-year students work as paid Teaching Fellows. Rates can change each year, so students should confirm current costs with Harvard Griffin GSAS.

View Harvard Griffin GSAS cost of attendance

Admission Requirements

Applicants need a bachelor’s degree or its equivalent. Applicants from a wide range of undergraduate disciplines are welcome.

SEAS does not set a GPA cutoff. Review is holistic, and the committee looks for demonstrated capacity for advanced computational work.

There are no formal prerequisites, but successful applicants are expected to have:

working knowledge of calculus, linear algebra, and differential equations
familiarity with probability and statistical inference
fluency in at least one programming language, such as Python or R
an understanding of basic computer science concepts

Application Materials

Applicants submit:

Harvard Griffin GSAS online application and $105 application fee
transcripts, uploaded electronically
resume or CV
three letters of recommendation
statement of purpose of about 1,000 words
proof of English proficiency when required

The GRE is optional. Interviews are not part of the process. Applicants may apply to either Data Science or Computational Science and Engineering, not both.

Applicants who need to show English proficiency submit the TOEFL iBT or IELTS Academic. Harvard Griffin GSAS asks for an IELTS overall score of 7 with 6.5 in speaking, or a TOEFL iBT overall score of 5 with 4.5 in speaking on the scale used since January 2026.

Application Deadlines

Entry Term	Deadline
Fall 2027	December 1, 2026, 5:00 p.m.

There is one deadline each year, with no rolling admission and no spring entry. Late applications are not accepted. Decisions are posted by mid-February.

Program Format

The SM in Data Science is taught in person, full time, during the day. Students take four 4-credit courses each semester and must be on campus for at least three semesters. There is no online option and no part-time entry.

SEAS has buildings in Cambridge and in the Allston neighborhood of Boston. Some students stay for a fourth semester to take more courses or complete a thesis, and the degree must be finished within four semesters.

The program is STEM designated. Admission offers are rarely deferred, and SEAS does not grant deferrals for employment or financing reasons.

Who Is This Program Best For?

Harvard’s SM in Data Science is a strong option for students with a solid mathematics and programming background who want a short, full-time degree that combines statistics and computer science.

It may be especially appealing to students interested in:

data science and machine learning
statistical inference and modeling
high performance and practical computing for data
ethics and critical thinking in data science
electives across Harvard schools and MIT
a capstone project or an optional research thesis

The degree is a data science master’s first. The machine learning course is one choice within a three-of-four requirement, and deep learning, natural language processing, and computer vision depend on elective selection. The program is on campus and full time only, admits once a year, and comes with high tuition and no Harvard financial aid. Students who want a degree built entirely around AI, online or part-time study, or lower costs may want to compare other Massachusetts schools.

Official Harvard University Program Links

Students can also compare Harvard University with other master’s programs in machine learning, artificial intelligence, and related fields offered in Massachusetts.

Boston College – Chestnut Hill
Boston University – Boston
Brandeis University – Waltham
Clark University – Worcester
Fitchburg State University – Fitchburg
Northeastern University – Boston
Tufts University – Medford
University of Massachusetts Amherst – Amherst
University of Massachusetts Lowell – Lowell
Wentworth Institute of Technology – Boston
Worcester Polytechnic Institute – Worcester

View All Machine Learning Programs in Massachusetts