MSc Data Science

What this degree is

MSc Data Science is a postgraduate degree at the intersection of statistics, machine learning, and data engineering. It trains students to extract knowledge from data — to build models, design analyses, interpret results, and communicate findings — at a level of methodological depth and computational sophistication that goes beyond what a three-year undergraduate programme can provide.

The degree is newer than MSc Mathematics or MSc Statistics. Most MSc Data Science programmes globally were established after 2012, with significant growth from 2016 onward as demand for data professionals outstripped supply and universities responded by creating dedicated postgraduate pathways. This youth has consequences: there is more variation in what the degree covers, how it is structured, and what it prepares graduates for than in more established degrees. A student considering MSc Data Science programmes needs to look carefully at the actual curriculum — the course list and the research orientation — not just the programme name.

The degree spans a spectrum from programmes that are primarily computational (machine learning, software systems, data engineering) to programmes that are primarily statistical (advanced inference, Bayesian methods, statistical modelling) to programmes that genuinely integrate both. The best programmes combine rigorous statistical foundations with practical machine learning implementation and real-world project experience.

In India, MSc Data Science is a growing offering at IITs, IIITs, and private research universities. The University of Edinburgh, LSE, Manchester, and UCL offer well-regarded one-year programmes in the UK. In the US, Columbia, Carnegie Mellon, Berkeley, and NYU offer competitive and well-resourced programmes at high cost. The duration varies significantly: one year in the UK, twelve to twenty months in most US programmes, and two years in India.

How MSc Data Science differs from BSc Data Science

The undergraduate BSc Data Science prepares students to apply data science tools and methods — Python, machine learning libraries, statistical techniques, visualisation — to structured problems. The curriculum is largely procedural in the first two years, building vocabulary and technique.

The MSc takes that foundation in several directions that the BSc cannot reach:

Research methods and original contribution. All MSc Data Science programmes include a substantial research project, dissertation, or thesis. Students are expected to make an original contribution — solving a novel problem, applying methods in a new domain, or developing a variant of a technique. This capacity for independent research is the primary distinguishing output of a postgraduate degree.

Mathematical and statistical depth. MSc programmes require — and teach — the mathematical foundations that are treated superficially at BSc level. Advanced machine learning theory (PAC learning, VC dimension, generalisation bounds), Bayesian inference, probabilistic graphical models, and optimisation theory all require mathematical maturity that most BSc programmes do not develop fully.

Advanced methods. Topics such as deep learning theory, reinforcement learning, causal inference, natural language processing at the research level, and Bayesian deep learning appear in MSc programmes. These appear at BSc level at best as applied survey topics; at MSc level they are studied with theoretical depth.

Domain application depth. MSc students typically work with real research or industry partners on substantive problems over an extended period. The Edinburgh MSc Data Science dissertation, for example, involves a full research project with academic or industry supervision. LSE’s capstone project runs from November to August — nearly ten months — with an actual company presenting a data problem.

A BSc Data Science graduate entering an MSc can expect the first few weeks of coursework to challenge their mathematical foundations and quickly move beyond the applied techniques they learned as undergraduates.

How MSc Data Science differs from MSc Statistics

The overlap between MSc Data Science and MSc Statistics is substantial — both deal with probability, statistical modelling, machine learning, and data analysis — but the degrees have different centres of gravity.

MSc Statistics places the theory of statistical inference at its centre: estimation theory, hypothesis testing, the mathematical foundations of probability, and the formal methodology of drawing conclusions from data. A statistics programme asks: what can we validly conclude from these observations, and why? The programme standards are set by statistical rigour — the correctness and efficiency of inferential procedures.

MSc Data Science places the working data pipeline at its centre: data collection and wrangling, feature engineering, model training and evaluation, deployment considerations, and scale. The question asked more often is: does this model perform well? A data science programme includes statistical methods but integrates them with computing, software, and systems thinking that statistics programmes typically do not.

Concretely: an MSc Statistics student at ISI or Edinburgh will spend significant time on the Neyman-Pearson Lemma, sufficiency, completeness, and the theoretical properties of estimators. An MSc Data Science student at Edinburgh (School of Informatics) or LSE will spend more time on machine learning algorithms, managing and visualising large datasets, and building systems that produce predictions at scale. Both will learn regression — but the statistical student will spend time on the theory of ordinary least squares and the Gauss-Markov theorem, while the data science student will spend more time on regularisation, cross-validation, and model selection in practice.

For students primarily interested in the mathematical foundations of inference, MSc Statistics is the right choice. For students who want to work at the intersection of statistics and software systems — building models, deploying solutions, working with databases and cloud infrastructure alongside statistical methodology — MSc Data Science is better aligned.

How MSc Data Science differs from MTech AI and Data Science

This distinction matters particularly for Indian students choosing between postgraduate programmes.

MTech AI/Data Science (offered at IITs and other technical universities) is an engineering degree — it sits in the engineering faculty, requires GATE (not JAM) for admission to most IIT programmes, and is structured around systems engineering alongside data science and AI methods. The MTech typically takes two years and involves a full research thesis. The engineering framing means students spend time on systems design, scalability, infrastructure, and production concerns that are less central in MSc programmes.

MSc Data Science is an academic science degree — it sits in a science, mathematics, informatics, or statistics faculty. The research orientation is toward advancing knowledge and methods, not toward engineering production systems. The MSc is better positioned for students planning PhD programmes in data science, ML, or related fields; it is also the standard credential recognised by international universities and research institutions.

In practice, both pathways lead to competitive industry careers. The relevant distinctions are:

• For PhD in data science or ML research: MSc is the natural stepping stone in India; MTech is also accepted by Indian PhD programmes
• For roles in industry AI engineering (large-scale ML systems, MLOps, infrastructure): MTech provides a more systems-oriented foundation
• For international career recognition and global study application: MSc from a recognised institution is more universally portable

Students with a BTech or BEng background may find the MTech more natural; students with a BSc in Mathematics, Statistics, or Computer Science typically enter MSc Data Science programmes more comfortably.

What students actually study

Foundation and core methods

Machine learning. Supervised learning (linear and logistic regression, decision trees, random forests, support vector machines, neural networks), unsupervised learning (clustering, dimensionality reduction, representation learning), reinforcement learning (policy gradient methods, Q-learning). At the MSc level, the theory — PAC learning, bias-variance tradeoff, VC dimension, generalisation bounds — is covered alongside the practical implementation. Edinburgh’s MSc Data Science (School of Informatics) covers Applied Machine Learning as an optional component; LSE’s programme includes Machine Learning and Data Mining as a compulsory course.

Probabilistic and statistical foundations. Probability theory, Bayesian inference (prior and posterior, MCMC methods, variational inference), probabilistic graphical models (Bayesian networks, Markov random fields), statistical testing, and regression theory. Carnegie Mellon’s MSML requires Probabilistic Graphical Models and Probability and Mathematical Statistics as core courses. These are the statistical foundations that serious ML research requires.

Data management and engineering. Relational databases, SQL, data wrangling and cleaning, working with unstructured data, data pipeline design, and (in more engineering-oriented programmes) distributed computing and cloud data infrastructure. LSE’s MSc Data Science includes Managing and Visualising Data as a compulsory course.

Statistical modelling. Generalised linear models, hierarchical models, time series analysis, spatial statistics, survival analysis, causal inference. These are the applied statistical methods used in research and in data-intensive industry.

Optimisation. Mathematical optimisation — gradient descent, stochastic gradient descent, convex optimisation, Newton’s method, constrained optimisation. Carnegie Mellon’s MSML includes Optimization for Machine Learning as a core requirement. Optimisation is the computational engine of machine learning.

Deep learning and neural networks. Feedforward networks, convolutional neural networks, recurrent neural networks, transformers, generative adversarial networks, large language models. At MSc level, both architecture and theory are covered.

Natural language processing. Text classification, sequence modelling, language models, named entity recognition, machine translation. Edinburgh’s Informatics-delivered MSc Data Science has strong NLP components; it sits within one of the UK’s leading NLP research groups.

Research methods and dissertation. All MSc Data Science programmes include a substantial original project. At Edinburgh (Informatics), this is a supervised research dissertation. At LSE, it is a capstone project running nearly ten months with an external industry partner. The project is typically the primary assessment of a student’s capacity for independent work.

Elective and specialisation areas

Depending on the programme, students may specialise in: computer vision; natural language processing; Bayesian machine learning; data-driven finance; healthcare and biomedical data; social data science and ethics; high-performance computing; or operations research and optimisation.

Typical curriculum structures

University of Edinburgh — MSc Data Science (School of Informatics):

Delivered by the School of Informatics, ranked 24th globally for Data Science and AI by QS World University Rankings 2025. The programme runs for one year and requires a substantial mathematics background: the equivalent of 60 SCQF credits (30 ECTS) covering calculus, linear algebra, discrete mathematics, and probability. Programming experience equivalent to an introductory course is required. Entry typically requires a first-class degree; offers are competitive. The programme covers machine learning, probabilistic and statistical methods, data management, and NLP. Students complete a supervised research dissertation.

LSE — MSc Data Science:

One-year programme in the LSE Department of Statistics. Four compulsory courses: ST443 Machine Learning and Data Mining, ST445 Managing and Visualising Data, ST447 Data Analysis and Statistical Methods, and ST498 Capstone Project (November to August — nearly ten months with an industry partner). Optional courses to the value of 1.5 units from a list including Deep Learning, Reinforcement Learning, Bayesian Machine Learning, Graph Data Analytics, Time Series, and Financial Statistics. Entry: UK 2:1 or equivalent in a relevant discipline with substantial mathematics. LSE’s location in London’s financial sector makes it particularly well-positioned for finance and policy data science roles.

Carnegie Mellon — MSML (Master of Science in Machine Learning):

Offered by CMU’s School of Computer Science, the MSML is one of the most prestigious and competitive data science/ML postgraduate degrees globally. Six core courses are required: Introduction to Machine Learning (or Advanced Introduction), Deep Reinforcement Learning or Advanced Deep Learning, Probabilistic Graphical Models, Machine Learning in Practice, Optimization for Machine Learning, and Probability and Mathematical Statistics (or Intermediate Statistics). Three electives are chosen from a list spanning advanced ML theory, NLP, computer vision, and domain applications. A practicum (industry project) is required. The programme takes three to four semesters. Highly selective; strong mathematical and programming background expected.

IIIT Hyderabad — Online MSc Data Science (DFL programme):

IIIT Hyderabad offers an online MSc in Data Science through its Distance and Flexible Learning (DFL) division. The programme runs over eight terms (two years, self-paced). Entry: Bachelor’s degree in basic sciences, technology, engineering, social sciences, commerce, or business; prerequisites include basic programming and mathematics (calculus, linear algebra, probability theory). The programme combines deep research foundations with real-world applications.

Shiv Nadar University — MTech Data Science and AI:

Shiv Nadar University offers an MTech in Data Science and Artificial Intelligence (not an MSc) through its School of Engineering. This is an engineering pathway. Eligibility: BTech (CSE, IT, ECE), MCA, or MSc in CS/IT/Electronics/Mathematics/Statistics; admission through written test and interview covering linear algebra, probability and statistics, calculus, discrete mathematics, and programming. Noted here as a reference for the MTech vs MSc distinction.

IIT Madras and University of Birmingham — Joint MSc Data Science and AI:

IIT Madras and the University of Birmingham offer an 18-month joint master’s programme in Data Science and Artificial Intelligence. The programme offers two tracks: one that allows students to complete a 12-month stay at the University of Birmingham (with Birmingham as the extended UK experience track) and one that returns to IIT Madras for the research project and final electives. The University of Birmingham is a partner in the Alan Turing Institute. The programme produces a joint degree recognised by both institutions.

Significant programme variation

MSc Data Science programmes vary more than most postgraduate degrees in what they actually deliver. Students should examine:

Who delivers the programme — informatics vs statistics vs management?

A programme housed in a computer science or informatics faculty (Edinburgh MSc Data Science — Informatics; Carnegie Mellon MSML) will emphasise algorithms, systems, and ML theory. A programme in a statistics department (LSE MSc Data Science — Statistics; Edinburgh MSc Statistics with Data Science — Mathematics) will emphasise statistical inference, modelling, and Bayesian methods. A programme in a business or management school (various) will emphasise applications, business analytics, and domain-specific methods with less mathematical depth.

Is programming a prerequisite or a core skill?

Serious MSc Data Science programmes at Edinburgh (Informatics), Carnegie Mellon, and LSE expect students who already know how to program. The degree goes beyond teaching Python syntax — it assumes computational fluency. If a programme’s first semester covers “introduction to Python,” it is likely not a research-level programme.

What does the dissertation involve?

A research dissertation (with a faculty supervisor and original research questions) is a different kind of output from an industry project or a coursework-heavy programme with a short capstone. Both have value, but they serve different purposes: research dissertations prepare students for PhD study and research careers; capstone projects prepare students for industry data science roles.

Skills this degree builds

Machine learning implementation and theory. Building, training, evaluating, and interpreting ML models from both a theoretical and practical perspective. Understanding when a model generalises and when it doesn’t.

Statistical reasoning applied to ML. Bayesian reasoning, uncertainty quantification, causal inference, experimental design — the statistical tools that allow data scientists to do more than just run models.

Data pipeline engineering. Managing, cleaning, transforming, and querying large datasets using Python, SQL, and data engineering tools. Understanding data provenance and quality.

Scientific communication. Writing research reports, presenting findings to technical and non-technical audiences, defending methodological choices. The dissertation or capstone project is specifically designed to develop this.

Research independence. Identifying a problem, reviewing relevant literature, choosing methods, executing analysis, interpreting results, and drawing conclusions. The postgraduate project is the primary vehicle for developing this.

Domain application. Through electives and projects, students develop depth in one or more application areas — finance, healthcare, NLP, computer vision, social science — that positions them for sector-specific roles.

Who should consider this degree

MSc Data Science is well-suited to students who:

• Have a bachelor’s degree in mathematics, statistics, computer science, or engineering and want rigorous postgraduate training in ML and statistical data science
• Are considering a research career in ML, data science, or related fields and want a stepping stone to PhD study
• Are working professionals (particularly relevant for online and part-time options) who want to upgrade statistical and ML skills to a research level
• Want a globally recognised credential — the UK MSc from Edinburgh, Imperial, or UCL, or a US MSc from CMU or Columbia — for international career mobility
• Are genuinely interested in both the mathematical foundations of ML and the practical engineering of data systems

It is not the right fit if:

• Your primary interest is in pure statistical theory and inference — MSc Statistics is better positioned for that
• Your interest is in the pure mathematics underlying ML (measure theory, functional analysis, algebraic structures) without the applied pipeline — MSc Mathematics is more appropriate
• You want primarily a professional credential with limited mathematical depth — industry-focused data analytics diplomas or MBA with analytics may be more efficient

Admissions and eligibility patterns

Common entrance routes

Route	Details
GRE	Required for most US MSc Data Science programmes (Columbia MSDS, Carnegie Mellon MSML, NYU, Berkeley MIDS). Most UK and Indian programmes do not require GRE, though a strong quantitative GRE score strengthens applications
College-specific (India)	IIT/IIIT programmes typically require GATE or an institutional written test. Shiv Nadar University MTech DS&AI: written test + interview covering linear algebra, probability, calculus, discrete mathematics, programming. IIIT Hyderabad online MSc: application review by admissions committee with prerequisites in programming and mathematics
College-specific (UK/Global)	Edinburgh, LSE, and most UK programmes: undergraduate transcript, personal statement, references. Edinburgh Informatics MSc Data Science explicitly requires first-class or equivalent for a competitive offer. IIT Madras – Birmingham joint MSc: separate application process to both institutions

Eligibility patterns:

Most MSc Data Science programmes require a bachelor’s degree in a quantitative discipline. Edinburgh Informatics requires mathematics equivalent to 60 SCQF credits (30 ECTS) in calculus, linear algebra, discrete mathematics, and probability, plus programming experience. LSE requires an upper second-class or equivalent with substantial mathematics. CMU’s MSML requires strong analytical skills and aptitude for mathematics, statistics, and programming — in practice, most admitted students have strong undergraduate records in CS, mathematics, or statistics.

Indian programmes typically accept graduates from BTech (CSE, ECE, IT), BSc (Mathematics, Statistics, Computer Science), or MCA backgrounds. Some programmes accept graduates from any quantitative field.

A note on GATE vs JAM:

MTech Data Science programmes at IITs typically use GATE scores. MSc Data Science programmes (where they exist as a distinct MSc, not MTech) may use IIT JAM or have their own entrance processes. Students should verify the specific admission gateway for each programme, as the MSc/MTech distinction affects both the entry route and the programme character.

India vs global degree structure

India

MSc Data Science as a distinct programme is still developing in India. Most IITs offer MTech (not MSc) in AI/Data Science at the postgraduate level, which is an engineering degree with different faculty affiliation, different entry routes (GATE), and a different research tradition. Some IITs and IIITs offer MSc Data Science programmes, particularly through online/distance learning divisions.

IIIT Hyderabad has offered an online MSc in Data Science through its Distance and Flexible Learning division, targeting students who want a research-grounded data science qualification with flexibility. The on-campus postgraduate data science offering at IIIT Hyderabad (and IIT Hyderabad, which has a department-specific MTech in Data Science for industry professionals) is primarily MTech-structured.

IIT Madras is notable for its joint MSc in Data Science and AI with the University of Birmingham — an 18-month programme producing a genuinely international credential from both institutions. This is distinct from IIT Madras’s well-known online BSc programme and from its BTech/MTech on-campus programmes.

Shiv Nadar University and several private research universities offer MSc or MTech programmes in data science and AI, with varying levels of research orientation. Shiv Nadar University has been one of the earlier private institutions to develop postgraduate data science offerings in India.

The general picture in India: serious quantitative postgraduate training in data science is available at IITs and IIITs either as MTech (engineering route, GATE) or through emerging MSc structures. The research-oriented MSc Data Science as a standalone programme comparable to Edinburgh or LSE is not yet widely institutionalised in India’s on-campus university system, though demand is clearly driving rapid development.

UK (one year)

UK MSc Data Science programmes run for one year (typically September to September). They are among the most sought-after postgraduate qualifications for data science internationally. Key programmes:

University of Edinburgh — MSc Data Science (Informatics): Delivered by the School of Informatics, which has research strengths in NLP, machine learning, computer vision, and speech. The programme ranks 24th globally for Data Science and AI (QS 2025). One year; strong mathematical prerequisites; competitive admissions (typical offer requires first-class equivalent). Dissertation with Informatics faculty supervision.

LSE — MSc Data Science: Statistics-department-housed programme with a distinctive emphasis on statistical foundations of data science. London location creates access to the finance, consulting, and policy sectors. Ten-month capstone project with industry partner. The statistical rigour is higher than many data science programmes; the engineering component is more moderate.

University of Edinburgh — MSc Statistics with Data Science: An alternative to the Informatics MSc Data Science for students who want more statistical depth. Delivered by the School of Mathematics; RSS-accredited. Compulsory Bayesian methods, generalised regression, statistical programming. More statistical theory than the Informatics programme; less systems/NLP focus.

Imperial College London, UCL, Manchester, Bristol, Warwick: All offer MSc Data Science programmes of varying emphases. Imperial’s Computing faculty programme leans toward ML engineering; UCL’s (in the Department of Statistical Science) balances statistics and ML; Manchester has strong industry links.

United States (12–20 months)

US programmes are highly resourced and internationally prestigious, but significantly more expensive than UK programmes. Scholarships and funding for master’s (as opposed to PhD) students are limited.

Carnegie Mellon MSML (School of Computer Science): One of the most prestigious ML master’s degrees globally. Six core courses build deep ML theory foundations; electives span the full ML research frontier. Three to four semesters; practicum required. Highly selective.

Columbia University MSDS (Data Science Institute): Interdisciplinary; combines statistics, computer science, and domain application. Strong New York tech industry links. Core includes machine learning, probability and statistics, algorithms, and exploratory data analysis.

UC Berkeley MIDS (School of Information): Emphasises applied data science in a research context. Strong links to Berkeley’s research centres in data science and AI.

NYU Center for Data Science — MS in Data Science: Interdisciplinary programme drawing on strengths in mathematics (Courant Institute), computer science, and domain sciences.

US programme durations range from 12 months (accelerated) to 20 months. Total cost (tuition plus living) is substantially higher than UK programmes; students should research funding options and post-graduation employment prospects carefully before committing.

A note on duration

A student comparing programmes across countries will encounter significant variation in duration: one year (UK), twelve to twenty months (US, IIT Madras-Birmingham), two years (India on-campus, some European programmes). Longer programmes offer more research depth and more time to develop skills; shorter programmes allow faster re-entry to the workforce. UK one-year programmes are legitimate, well-regarded qualifications, not compressed versions of longer degrees — they are designed as intensive immersive programmes with that duration in mind.

Careers after this degree

Data scientist / machine learning engineer. The primary industry destination. Roles in technology companies, financial services, consulting, healthcare, and media. The MSc provides the methodological depth that differentiates a data scientist who can justify their modelling choices from one who can only execute them. Industry entry-level roles: INR 12-25 LPA in India (approximate, varying by employer and city); significantly higher internationally.

Research scientist. At technology company research labs (Google DeepMind, Meta AI, Microsoft Research, AWS), academic-adjacent institutes, and national research organisations. These roles typically require strong mathematical foundations and the ability to produce novel research. The dissertation experience from an MSc is directly relevant.

Quantitative analyst (quant) or risk modeller. In finance, combining ML and statistical methods for trading, risk management, credit scoring, and derivatives pricing. LSE’s London location and finance-oriented optional courses make its graduates particularly competitive for these roles.

Data engineer / ML infrastructure engineer. Building and maintaining data pipelines, ML platforms, and data infrastructure. More systems-engineering oriented; the MSc provides the understanding of what the infrastructure needs to support.

AI/ML product manager. Managing AI-powered products; requires understanding of ML capabilities and limitations at a level the MSc provides.

PhD in data science, ML, or statistics. The MSc is the natural stepping stone to doctoral study. Strong research dissertations from Edinburgh (Informatics), LSE, or CMU open doors to top PhD programmes globally. For Indian students, MSc or MTech from IITs and IIITs similarly qualifies for PhD applications.

Academic faculty (via PhD). Eventually, graduates of the MSc who complete strong PhD programmes may enter faculty positions at universities, where they teach and research in data science, ML, and related areas.

Higher study and progression pathways

• PhD in Machine Learning / Data Science / Informatics — Edinburgh, Cambridge, Oxford, CMU, MIT, Berkeley, Stanford internationally; IIT, IISc, IIIT, ISI in India
• PhD in Statistics — for graduates who want to develop the statistical theory side more deeply
• PhD in Computer Science — for graduates whose primary interest is in systems and algorithms alongside ML
• Second MSc — some graduates pursue a further specialised MSc (MSc Biostatistics, MSc Financial Mathematics, MSc Actuarial Science) to deepen a specific application area
• Research fellowships — NBHM (mathematics), DST-Inspire (science), and national research fellowships support transition to doctoral study

Indian institutional examples

IIIT Hyderabad — Online MSc Data Science: Combines deep research foundations with real-world applications. Eight-term, two-year online programme open to graduates from any STEM, social science, or business background with programming and mathematics prerequisites. Total fee approximately INR 4 lakhs (plus GST). Programme is self-paced within the eight-term structure.

IIT Madras and University of Birmingham — Joint MSc Data Science and AI: An 18-month programme with study at both institutions. Includes an industrial placement opportunity. The University of Birmingham is a partner in the UK’s Alan Turing Institute. Produces a joint degree with recognition from both IIT Madras and the University of Birmingham. See the official joint programme page.

Shiv Nadar University — offers an MTech in Data Science and Artificial Intelligence through its engineering faculty. The programme covers core and elective courses with a research thesis component. Admission through written test and interview; eligibility includes BTech, MCA, or MSc in relevant disciplines.

Note: Several Indian institutions, including IIT Hyderabad’s CSE Department, offer an MTech in Data Science specifically for industry professionals with three or more years of work experience — this is not a fresh-graduate MSc. Students should confirm whether a programme they are considering accepts fresh graduates or requires industry experience.

International institutional examples

University of Edinburgh — MSc Data Science (School of Informatics): The UK’s highest-ranked data science and AI programme at postgraduate level (QS 24th globally, 2025). One year; delivered by Informatics faculty with research strengths in NLP, ML, and computer vision. Requires first-class equivalent for competitive offers; strong mathematics and programming prerequisites. Dissertation with faculty supervision. See official Edinburgh Data Science MSc information via Collegedunia for entry requirements.

LSE — MSc Data Science: Statistics-grounded programme with compulsory courses in machine learning and data mining, managing and visualising data, and statistical methods. Ten-month capstone project with an external industry partner. Optional courses in deep learning, reinforcement learning, Bayesian ML, graph analytics, time series, and financial statistics. Strong finance-sector placement given LSE’s London location. See the LSE MSc Data Science programme page.

Carnegie Mellon MSML: The most prestigious standalone ML master’s degree globally. Core: Introduction to ML, Deep Reinforcement Learning or Advanced Deep Learning, Probabilistic Graphical Models, ML in Practice, Optimization for ML, and Probability and Mathematical Statistics. Electives span advanced ML theory, NLP, computer vision, and domain applications. Practicum required. STEM-designated. See the CMI MSML curriculum page.

University of Edinburgh — MSc Statistics with Data Science: The alternative programme for students wanting more statistical theory alongside data science methods. Delivered by the School of Mathematics; RSS-accredited. Compulsory Bayesian Data Analysis, Bayesian Theory, Generalised Regression Models, and Extended Statistical Programming. Optional applied machine learning, biostatistics, causal inference, and text technologies. Consultancy-style dissertation. Entry: UK 2:1 with substantial maths including calculus, linear algebra, probability, and statistical theory. See the official Edinburgh Statistics with Data Science programme page.

Related degrees and next reads

• BSc Data Science — the undergraduate precursor; MSc goes substantially deeper in theory and research methods
• MSc Statistics — statistical theory and inference emphasis vs data engineering and ML pipeline emphasis
• MSc Mathematics — mathematical depth without the applied ML/data pipeline orientation
• BSc Statistics — undergraduate statistics as an alternative entry background
• BSc Mathematics — mathematical preparation relevant for MSc Data Science applications
• MSc Economics — for students interested in quantitative economic analysis; shares overlap with social data science electives
• BTech AI and Data Science — the undergraduate engineering route to AI/data science

Sources Used

• University of Edinburgh MSc Data Science — Collegedunia — programme details, QS 24th global ranking, entry requirements for Indian students (55% or 7.0 GPA from listed universities)
• University of Edinburgh MSc Data Science entry requirements — Whatuni — mathematics prerequisite detail (60 SCQF credits covering calculus, linear algebra, discrete maths, probability), programming experience requirement
• University of Edinburgh MSc Statistics with Data Science — official programme page — compulsory courses, RSS accreditation, consultancy dissertation structure, entry requirements
• LSE MSc Data Science — official LSE programme page — compulsory courses (ST443, ST445, ST447, ST498 Capstone), optional courses list, entry requirements
• LSE MSc Data Science programme structure document (PDF) — four compulsory units, 1.5 optional units, capstone project November–August
• Carnegie Mellon MSML curriculum — official CMI page — six core courses, three electives, practicum, STEM designation
• Carnegie Mellon MSML Handbook AY25-26 (official PDF) — programme variants (standard, applied, advanced study), completion requirements
• IIIT Hyderabad Online MSc Data Science — DFL programme page — eight-term two-year programme, eligibility, fee structure (INR 4 lakhs + GST), prerequisites
• IIT Madras–University of Birmingham Joint MSc Data Science and AI — official programme page — 18-month duration, two tracks (Birmingham extended, IIT Madras return), Alan Turing Institute partnership
• IIT Hyderabad MTech Data Science (industry professionals) — official brochure PDF — confirmed as industry-professionals-only programme; relevant to MSc vs MTech distinction
• Shiv Nadar University MTech Data Science and AI — official programme announcement (PDF) — eligibility (BTech CSE/ECE/IT, MCA, MSc CS/IT/Electronics/Maths/Statistics), written test syllabus, engineering degree confirmed
• BSc Data Science at IIT Madras online — TalentSprint reference — IIT Madras data science ecosystem; 15 faculty, executive certification noted as distinct from on-campus MSc