Pt I: Hyperdimensional Computing (HDC) with Peter Sutor (Interview) скачать в хорошем качестве

Pt I: Hyperdimensional Computing (HDC) with Peter Sutor (Interview)

Hyperdimensional computing (HDC) involves the encoding of data into very high dimensional vectors of binary variables, mimicking the redundancy of brain function - a sharp contrast to the lower dimensional embeddings used in most machine learning algorithms. In this interview, we meet Peter Sutor, a research at Simuli who is finishing his doctorate at the University of Maryland where he has written extensively on this topic. Peter's paper "Learning sensorimotor control with neuromorphic sensors: Toward hyperdimensional active perception": https://www.researchgate.net/publicat... Pentti Kanerva's seminal paper on HDC: https://rctn.org/vs265/kanerva09-hype... 0:00 Introduction 5:18 Introduction to HDC 14:29 Basic operations on hyperdimensional vectors 18:10 Inference with HDC 26:30 Hyperdimensional inference layer 34:12 Efficient vision learning with HDC 42:30 Discussion on efficiencies of HDC observed in experiments 45:30 Future directions About the channel: The Neuro Symbolic Channel provides the tutorials, courses, and research results on one of the most exciting areas in artificial intelligence and machine learning. With content originally from the AI course taught at Arizona State University, this channel brings you the latest at the intersection of symbolic methods (e.g., logic programming) and deep learning. Learn about the latest algorithms, Python packages, and progress toward larger goals such as artificial general intelligence (AGI).