Hyperdimensional computing: Difference between revisions

Revision as of 13:23, 27 April 2023

Revision as of 19:00, 28 April 2023

Line 6:

== Process ==

Data is mapped from the input space to sparse HD space under an encoding function φ : X → H. HD representations are stored in data structures that are subject to corruption by noise/hardware failures. Noisy/corrupted HD representations can still serve as input for learning, classificationgclassification, etc. They can also be decoded to recover the input data. H is typically restricted to range-limited integers (-v-v)<ref name=":1">{{Cite journal |last=Thomas |first=Anthony |last2=Dasgupta |first2=Sanjoy |last3=Rosing |first3=Tajana |date=2021-10-05 |title=A Theoretical Perspective on Hyperdimensional Computing |url=https://redwood.berkeley.edu/wp-content/uploads/2021/08/Thomas2021.pdf |journal=Journal of Artificial Intelligence Research |language=en |volume=72 |pages=215–249 |doi=10.1613/jair.1.12664 |issn=1076-9757}}</ref>

This is analogous to the learning process conducted by [[fruit flies]] olfactory system. The input is a roughly 50-dimensional vector corresponding to odor receptor neuron types. The HD representation uses ~2,000-dimensions.<ref name=":1" />