Difference between revisions of "Improving Resiliency of Hyperdimensional Computing"
From iis-projects
(Created page with "Category:Digital Category:Human Intranet Category:Completed Category:Semester Thesis Category:2019 Category:Herschmi Hyperdimensional (HD) computing [1...") |
|||
Line 1: | Line 1: | ||
[[Category:Digital]] [[Category:Human Intranet]] [[Category:Completed]] [[Category:Semester Thesis]] [[Category:2019]] | [[Category:Digital]] [[Category:Human Intranet]] [[Category:Completed]] [[Category:Semester Thesis]] [[Category:2019]] | ||
− | [[Category:Herschmi]] | + | [[Category:User:Herschmi]] |
Hyperdimensional (HD) computing [1] is a brain-inspired computing paradigm based on representing information with hypervectors (thousands of dimensions). Hypervectors are holographic and (pseudo)random with independent and identically distributed (i.i.d.) components. The most important aspect of HD computing, for hardware realization, is its robustness against noise and variations in the computing platforms [3]. Principles of HD computing allow to implement efficient machine learning models as well as universal computing, e.g., emulating finite state automata [4]. By its very nature, HD computing is extremely robust against failures, defects, variations, and noise, all of which are synonymous to ultra low energy computation on nanoscale fabrics [2]. In this project, your goal would be to improve algorithms based on HD computing to become more robust against failures. | Hyperdimensional (HD) computing [1] is a brain-inspired computing paradigm based on representing information with hypervectors (thousands of dimensions). Hypervectors are holographic and (pseudo)random with independent and identically distributed (i.i.d.) components. The most important aspect of HD computing, for hardware realization, is its robustness against noise and variations in the computing platforms [3]. Principles of HD computing allow to implement efficient machine learning models as well as universal computing, e.g., emulating finite state automata [4]. By its very nature, HD computing is extremely robust against failures, defects, variations, and noise, all of which are synonymous to ultra low energy computation on nanoscale fabrics [2]. In this project, your goal would be to improve algorithms based on HD computing to become more robust against failures. | ||
Revision as of 15:32, 29 October 2019
Hyperdimensional (HD) computing [1] is a brain-inspired computing paradigm based on representing information with hypervectors (thousands of dimensions). Hypervectors are holographic and (pseudo)random with independent and identically distributed (i.i.d.) components. The most important aspect of HD computing, for hardware realization, is its robustness against noise and variations in the computing platforms [3]. Principles of HD computing allow to implement efficient machine learning models as well as universal computing, e.g., emulating finite state automata [4]. By its very nature, HD computing is extremely robust against failures, defects, variations, and noise, all of which are synonymous to ultra low energy computation on nanoscale fabrics [2]. In this project, your goal would be to improve algorithms based on HD computing to become more robust against failures.
Status: Completed
Sara Sangalli
- Supervision: Michael Hersche, Abbas Rahimi
- Date: 5/2019
Prerequisites
- Machine Learning
- Matlab Programming
Character
- 40% Theory
- 60% Programming
Professor
Literature
- [1] Kanerva, Hyperdimensional Computing: An Introduction to Computing in Distributed Representation with High-Dimensional Random Vectors
- [2] Rahimi et al., High-Dimensional Computing as a Nanoscalable Paradigm
- [3] Rahimi et al., A robust and energy efficient classifier using brain-inspired hyperdimensional computing
- [4] Osipov et al., Associative synthesis of finite state automata model of a controlled object with hyperdimensional computing