Hardware Implementation of Limited-Precision Machine-Learning Algorithms for Edge Computing

Hardware Implementation of Limited-Precision Machine-Learning Algorithms for Edge Computing

Advisor/s

Mostafa Rahimi Azghadi

College or Research Centre

College of Science & Engineering

Summary of Project

Machine learning algorithms have shown unprecedented performance in many challenging engineering applications such as machine vision, speech processing, and autonomous learning. These algorithms though, require huge processing capability and consume a high amount of power, which are not available on edge devices in the Internet of Things. This project investigates, design, and implement machine-learning and neuromorphic learning algorithms in hardware to enable edge computing. The developed hardware could be first prototyped on FPGAs and then implemented as an electronic chip to carry out a learning task such as character recognition.

Key Words

machine-learning; neuromorphic computing; IoT; edge computing; FPGA; engineering

Would suit an applicant who

Has an interest in hardware and software design of machine learning algorithms

Updated: 08 Apr 2020