Duke Masters in Machine Learning Degree

Program Overview

Pratt School of Engineering

Duke University’s Machine Learning & Big Data master’s track is part of the Electrical and Computer Engineering graduate program and is designed for students seeking advanced training in data-driven technologies.

This program requires the completion of at least 10 graduate-level courses plus a seminar, totaling 30+ credits, with options for additional research and elective coursework.

Students may pursue this track through either the Master of Science (MS) or the Master of Engineering (MEng) degree paths.

Program Format

This program is campus-based in Durham, North Carolina. However, its flexible, individualized curriculum allows students to align their studies with personal career goals.

Full-time students typically complete the program in 16 to 24 months, depending on course load and thesis or internship involvement.

Curriculum and Specialization

The curriculum provides a strong foundation in programming, data structures, machine learning, statistics, and deep learning. Required coursework includes Mathematics for Machine Learning, Practical Machine Learning, Data Engineering, and Deep Neural Networks.

Students further customize their learning through elective options such as Natural Language Processing, Information Theory, Adversarial Machine Learning, and Signal Detection. An optional research component or thesis allows for deeper exploration of advanced topics.

Courwork includes the following:

ECE 551D Programming, Data Structures, and Algorithms in C++ introduces programming in C and C++ with emphasis on core data structures, abstract data types, and common algorithms. The course includes efficiency analysis using Big-O notation and introduces concurrent programming concepts. Students gain practical experience using UNIX development tools and defensive coding strategies.

ECE 590 P4ML Programming and Data Structures for Machine Learning provides instruction on programming techniques and data structures relevant to machine learning applications. Topics are tailored for machine learning contexts and often include implementation exercises.

ECE 590 MML Mathematics for Machine Learning focuses on mathematical foundations necessary for machine learning, with topics chosen to meet the needs of specialized student groups. Enrollment requires instructor approval.

ECE 590 PML Practical Machine Learning offers hands-on study of machine learning methods with practical application and discussion. Designed to support real-world problem-solving using ML techniques.

ECE 590 DE Data Engineering introduces data engineering techniques including data pipelines, storage systems, and data preprocessing, preparing students for ML and analytics workflows.

ECE 661 Computer Engineering ML & Deep Neural Networks explores deep learning techniques and engineering methods for optimizing neural networks. Students implement models in PyTorch to evaluate accuracy, size, and performance across computing platforms.

ECE 685D Introduction to Deep Learning covers deep neural networks from a mathematical and implementation perspective. Students complete programming assignments in Python and use PyTorch or TensorFlow to apply deep learning to real-world problems.

ECE 558 Advanced Computer Network examines advanced networking protocols, performance optimization, and emerging technologies for distributed systems.

ECE 565 Performance Optimization & Parallelism teaches strategies to improve computational performance through algorithmic optimization and parallel computing techniques.

ECE 585 Signal Detection & Extraction Theory provides a foundation in signal detection and estimation theory. Students learn methods for detecting known and random signals in noisy environments with applications in communication systems.

ECE 587 Information Theory introduces core information theory concepts such as entropy, mutual information, and data compression. Applications span communications, inference, and statistical modeling.

ECE 588 Image and Video Processing covers image compression, enhancement, and segmentation, with extensions to video and applications in medical imaging and computer vision. Students learn geometric tools and compressed sensing techniques.

ECE 662 ML Acceleration of Neuromorphic Computing explores hardware-efficient design of neural networks, covering GPU, FPGA, and ASIC-based implementations. Topics also include bio-inspired computing and neuromorphic systems.

ECE 663 Machine Learning in Adversarial Settings focuses on vulnerabilities in ML systems and methods for securing algorithms against adversarial attacks. Students study both threats and defenses.

ECE 681 Pattern Classification & Recognition Technology teaches recognition algorithms used in intelligent systems. Students apply pattern recognition to problems in healthcare, biometrics, and weather modeling.

ECE 682D Probabilistic Machine Learning introduces graphical models and probabilistic methods for machine learning. The course emphasizes parameter estimation, kernel methods, and structure learning.

ECE 684 Natural Language Processing explores computational methods for processing and analyzing text data. Applications include sentiment analysis, OCR, and AI-driven assistants, using Python-based toolkits.

ECE 687D Theory and Algorithms for Machine Learning presents a survey of classical ML algorithms and statistical learning theory. Topics include support vector machines, neural networks, ensemble methods, and unsupervised learning.

ECE 689 Advanced Topics in Deep Learning investigates current deep learning techniques such as physics-informed models and generative methods. Students implement and evaluate advanced models in practical assignments.

ECE 590 in AI Advanced Topics in ECE – AI-related allows customized study in artificial intelligence with advisor approval. Content varies based on emerging trends and student interests.

ECE 899 Indep. involves guided independent study in a chosen electrical engineering area. Requires approval from the director of graduate studies.

MEng 540 is a core requirement for Master of Engineering students focusing on engineering management and applied technical skills.

MEng 570 is a second core requirement that develops leadership and strategic decision-making for engineering professionals.

More information about the curriculum here:

Admission and Requirements

Applicants must have a strong background in mathematics, programming, and engineering or a related field. Those confident in their programming skills may apply for a waiver of the introductory programming course.

All students are required to attend the ECE 701S first-year seminar, which fosters professional development and research exposure.

Tuition

The total tuition cost is approximately $31,000 to $35,000, depending on the number of credits taken and program format (MS vs. MEng). Additional fees and living expenses are not included in this estimate.

More tuition information here: https://gradschool.duke.edu/financial-support/

Students may be eligible for research assistantships or financial support through Duke’s affiliated research centers.

Ideal Candidates and Career Outcomes

This program is ideal for students aiming to enter or advance in fields such as machine learning engineering, data science, algorithm development, or AI research. It also suits professionals seeking to pair technical mastery with real-world application, especially those interested in interdisciplinary innovation across technology and health, finance, or public policy sectors.

Graduates benefit from Duke’s global alumni network and are prepared for roles such as ML engineer, data scientist, or AI researcher.

This rigorous, research-driven program offers a strong foundation for those looking to lead in machine learning, with access to leading faculty, cutting-edge research centers like the Rhodes Information Initiative, and industry-relevant curriculum designed for high-impact careers.