Modern machine learning methods including deep learning have shown super-human performance in some applications. However, they face several challenges in algorithm and data limitations. In the last few years, Hyperdimensional Computing (HDC) has emerged as a promising paradigm to tackle the challenges faced by existing machine learning and deep learning. It is a brain-inspired emerging computing paradigm for representing and manipulating concepts and their meanings using fixed-size vector representations in a high-dimensional vector space. HDC has shown how these vector representations can be used to perform ‘brain-like’ neuromorphic cognitive processing.

The proposed project harnesses advancements from hyperdimensional computing, to develop light weight algorithms for efficiently performing cognitive processing for applications where computational resources and energy capacity are limited, such as at the network edge. It builds on our previous work sponsored by the U.S. Army Research Laboratory and the U.K. Ministry of Defence. In particular this project will develop an anomaly detection method based on the distinct and inherent features of hyperdimensional computing. It will implement the method using the “vsapy” (https://pypi.org/project/vsapy/), a Python package developed by our recent PhD student, or other HDC packages.

This project requires strong experience and skills in machine learning, Python programming and maths (linear algebra in particular). It is particularly suitable for students who wish to pursue PhD study in machine learning and data science after dissertation.

Related paper: R. Wang, F Kong, H Sudler, X Jiao (2021) "HDAD: Hyperdimensional Computing-based Anomaly Detection for Automotive Sensor Attacks," IEEE 27th Real-Time and Embedded Technology and Applications Symposium.