Interpolation coding: A representation for numbers in neural models

DOI10.1007/BF00354984MaRDI QIDQ580249zbMATH OpenOpenAlexWikidataFDO

Authors Dana H. Ballard

Publication date 1987

Published in Biological Cybernetics (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/bf00354984

zbMATH Keywords

computing hierarchies of invariantsencodingfunctional mappingsinterpolation codinglevels of abstractionneural modelsperceptionvalue units

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Artificial intelligence (68T99) Psychophysics and psychophysiology; perception (91E30) Mathematical psychology (91E99) Physiological, cellular and medical topics (92Cxx) Communication, information (94A99)

Review text

A central task of perception can be defined as one of computing hierarchies of invariants. One way of representing such invariants in intermediate levels of abstraction in this hierarchy is to use discrete units. These have been termed value units. A problem with such an encoding is that there has not been a good way to represent accurate numerical quantities using these units. This paper remedies the deficiency by describing a scheme that interpolates values between units representing fixed numerical quantities. The scheme has nice properties: it extends across functional mappings and it allows different sources of evidence to be combined.

Recommendations

• A new type of neural computation
• Population coding and the labeling problem: extrinsic versus intrinsic representations
• A competitive neural model of small number detection
• scientific article; zbMATH DE number 2013219
• Neural coding and decoding: communication channels and quantization

Cites work

• ``Neural computation of decisions in optimization problems
• scientific article; zbMATH DE number 3447323 (Why is no real title available?)
• Inherent Bias and Noise in the Hough Transform
• Neural networks and physical systems with emergent collective computational abilities
• Neurons with graded response have collective computational properties like those of two-state neurons
• On the Foundations of Relaxation Labeling Processes
• Optimization by simulated annealing
• Scene Labeling by Relaxation Operations
• Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images
• Synthesizing constraint expressions
• Use of the Hough transformation to detect lines and curves in pictures

Cited in

(3)

• Counterfactual state explanations for reinforcement learning agents via generative deep learning
• A perceptual-like population-coding mechanism of approximate numerical averaging
• A competitive neural model of small number detection

This page was built for publication: Interpolation coding: A representation for numbers in neural models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q580249)