Associative Memory: A System-Theoretical Approach

By Teuvo Kohonen. 1977

Subfields: Model Architectures, Neuroscience

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Model Architectures

In Neuroscience

Citations in the corpus

Neural networks and physical systems with emergent collective computational abilities. (1982, background): “Such matrices T,. have been used in theories of linear associative nets (15-19) to produce an output pattern from a paired input stimulus, S1 -* 01.”
Theory for the development of neuron selectivity: orientation specificity and binocular interaction in visual cortex (1982, background): “It is known that various algorithms related to Hebb’s principle of synaptic modification (Hebb, 1949) can account for the formation of associative and distributed memories (see, for example, Marr, 1969; Brindley, 1969; Anderson, 1970, 1972; Cooper, 1973; Kohonen, 1977).”
A Survey on Electric Power Demand Forecasting: Future Trends in Smart Grids, Microgrids and Smart Buildings (2014, methodology): “The same year, at the University of Helsinki, Teuvu Kohonen presented an enhancement of Anderson’s model, coming up with SOM (Kohonen, 1977 [41]).”
3 Learning distributed representations of concepts (2010, background): “In the componential approach each concept is simply a set offeatures and so a neural net can be made to implement a set of concepts by assigning a unit to each feature and setting the strengths of the connections between units so that each concept corresponds to a stable pattern of activity distribu”
An interactive activation model of context effects in letter perception: part 1.: an account of basic findings (1988, background): “James Anderson and his colleagues (Anderson , 1977; Anderson, Silverstein, Ritz, & Jones, 1977) and Kohonen and his colleagues (Kohonen, 1977) have developed a pattern recognition system which they call an associative memory system.”
MACE: Mass Concept Erasure in Diffusion Models (2024, methodology): “Drawing upon methods that view matrices as linear associative memories [1, 28], often used to edit knowledge embedded within neural networks [2, 4, 5, 17, 36, 37, 43], we formulate our objective function as follows: where λ 1 ∈ R + is a hyperparameter, e is the embedding of a word co-existing with…”
The mechanisms for pattern completion and pattern separation in the hippocampus (2013, background): “Correlations between patterns reduce the memory capacity of an autoassociation network (Marr, 1971; Kohonen, 1977, 1984; Kohonen et al., 1981; Rolls and Treves, 1998), and because storage capacity is at a premium in an episodic memory system, there are several mechanisms that reduce the…”
Pattern separation, completion, and categorisation in the hippocampus and neocortex. (2016, methodology): “2015 Published by Elsevier Inc. 53 66 67 68 69 70 71 72 73 74 75 76 77 78”
Optimal synaptic learning in non-linear associative memory (2010, background): “Neural associative memories have close relationships to Hebbian cell assemblies [1] and are widely used in neuro- science as models of neural computation for various brain structures, for example, neocortex [2], [3], [4], [5], [6], hippocampus [7], [8], perirhinal cortex [9], cerebellum [10], [11],…”
Serial Order: A Parallel Distributed Processing Approach (1997, background): “In networks with no hidden units, learning rules have been developed that allow any linearly separable function to be learned (Kohonen, 1977; Rosenblatt, 1962; Widrow & Hoff, 1960).”
Toward a modern theory of adaptive networks: expectation and prediction. (1981, background): “The most well-developed application of the Hebbian learning rule is its use in networks that implement associative information storage (e.g., Amari, 1977a, 1977b; Anderson et al., 1977; Kohonen , 1977; Nakano, 1972; Wigstrom, 1973).”
Self-organized formation of topologically correct feature maps (2004, background): “This is possible if the synaptic efficacy is mainly determined by one or more postsynaptic factors which are redistributed between the synaptic sites in proportion to their use (Kohonen, 1977).”
Synchronization and pattern separation in spiking associative memories and visual cortical areas (2004, methodology): “Diese Arbeit leistet hauptsächlich Beiträge zur Theorie der Neuronalen Assoziativspeicher [174, 204, 104, 129, 69, 105, 134, 29, 160, 171] und zur Theorie der Neuronalen Assemblies des Gehirns [64, 26, 130, 192, 133].”
A massively parallel architecture for a self-organizing neural pattern recognition machine (1988, methodology): “Unlike many alternative models [15-19], the present model can deal with arbitrary combinations of binary input patterns.”
Nonlinear neural networks: Principles, mechanisms, and architectures (1988, methodology): “The streams of research TABLE 1 Kohonen (1971) made his transition from linear algebra concepts such as the Moore-Penrose pseudoinverse to more biologically motivated studies which he has summarized in his influential books (Kohonen, 1977, 1984).”