Machine learning is a transformative force in technology, particularly in its application within embedded systems. These systems are specialized computer systems designed for specific tasks. Integrating machine learning into these systems, such as those found in self-driving cars, medical devices, and smart appliances, presents unique challenges. These include the necessity for algorithms to process data on the edge - directly where it's captured - to enhance response times and reduce bandwidth demands. This requirement brings forth challenges in accommodating machine learning within the limited resource capacities of embedded systems. It also necessitates a balance between accuracy and the constraints of hardware resources, including power consumption and latency. Addressing these challenges requires specialized knowledge in hardware-aware machine learning and co-design approaches. Consequently, the certificate aims to equip students with the skills to develop efficient, secure, and cost-effective machine learning solutions for embedded systems, focusing on optimal algorithm implementation, software/hardware co-design, federated learning, and enhanced security measures for embedded machine learning applications. The primary target audience includes graduates of programs in computer engineering or electrical engineering whose degree programs did not include machine learning coursework. A second target audience includes professional embedded software developers writing code for microcontrollers and hardware designers who do not have a background in machine learning.

Admissions

The graduate certificate is open to all students who hold baccalaureate degrees in scientific and engineering disciplines from accredited universities in accordance with Graduate Admission Requirements. Non-degree graduate students are welcome to apply. Current degree seeking students in the College of Engineering and Computing may request admission as a secondary certificate.

Policies

A cumulative GPA of 3.00 is required, and at most one course with a grade of C may be applied toward the certificate. 

For policies governing all graduate certificates, see AP.6.8 Requirements for Graduate Certificates.

Banner Code: EC-CERG-MLES

Certificate Requirements

Total credits: 12

This certificate may be pursued on a full-or part-time basis.

Required Courses

ECE 527Learning From Data3
ECE 554Machine Learning for Embedded Systems3
Total Credits6

Electives

Select two courses from the following:6
ECE 556Neuromorphic Computing
ECE 552Big Data Technologies
ECE 617Distributed and Federated Learning
ECE 618Hardware Accelerators for Machine Learning
ECE 651Advanced Learning From Data
ECE 653Machine Learning Security and Privacy
Total Credits6