Object Generation with Neural Networks (When Spurious Memories are Useful)

Ezhov, Alexandr A.; Vvedensky, V. L.

doi:10.1016/s0893-6080(96)00064-0

Cited by 14 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It has been shown [3] that this position in the table actually determines in many cases the different phenotypic properties of viruses. Note, that neural replicators are built on the basis of energy-minimizing Hop eld neural networks [9] with different neural thresholds, and the motifs of the replicators can be interpreted as prototypes, the meaning of which corresponds to the polythetic nature of pattern classi cation by the Hop eld network [13].…”

Section: Discussionmentioning

confidence: 99%

Can the natural system of viruses reconcile the current taxonomy with an alternative classification useful to clinicians?

Ezhov

2023

Preprint

View full text Add to dashboard Cite

In 2022, a group of basic and clinical virologists, bioinformaticians, and evolutionary and structural biologists met in Oxford, UK, to develop a consensus on methodologies used to classify viruses. They concluded that virus taxonomy, which is hierarchical and based on evolution, is only one of many possible ways to classify viruses. This taxonomy, while satisfying the four principles they set out, faces difficulties in coordinating with other classification systems useful to clinicians, infectious disease specialists, agronomists, etc. One example discussed is the grouping of different viral strains that cause different diseases into the species Enterovirus C. Here we show that the use of a previously proposed variant of a natural virus classification system based on the use of Neural Replicator Analysis can resolve this contradiction by establishing the fine structure of the Enterovirus C species, in which strains that cause different diseases are placed in several different cells of the binomial table of viruses. A key element in enabling this is the sophisticated preprocessing of the original viral genomes using neural replicators.

show abstract

Section: Discussionmentioning

confidence: 99%

Can the natural system of viruses reconcile the current taxonomy with an alternative classification useful to clinicians?

Ezhov

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Ezhov and Vvedensky (1996), among others have described a method of generating attractors within a neural network. Networks may contain several attractors that can be activated individually from one another, allowing the network to distinguish between different input node activation patterns.…”

Section: Neural Networkmentioning

confidence: 99%

The state of play in machine/environment interactions

Aitkenhead

McDonald

2006

Artif Intell Rev

View full text Add to dashboard Cite

Due to the breadth of the subject, it is no longer possible to provide a review of all of the work being carried out in the field of Artificial Intelligence. However, a more localised review of research taking place in the overlap between engineering, AI and psychology can be meaningfully performed. We show here that while there have been marked successes in the past few years, there is an identifiable set of 'classic' problems that remain to be solved, and which largely direct the work ongoing in this area. This review aims to discuss the directions being taken at the current time, in particular the developing and maturing possibilities provided by neural networks and evolutionary computation, and by the use of our knowledge of the mind in developing artificial agents capable of mimicking our abilities to interact with the environment.

show abstract

“…In particular, an analogous architecture is presented by Ezhov and Vvedensky. 11 The authors of the referred paper name every sub-network of their network as a single-class network. Every of these single-class networks is an attractor neural network which works according to a Hopfield network manner.…”

Section: An Outline Of the Assembly Neural Networkmentioning

confidence: 99%

Generalization of Features in the Assembly Neural Networks

Goltsev

Wunsch

2004

Int. J. Neur. Syst.

View full text Add to dashboard Cite

The purpose of the paper is an experimental study of the formation of class descriptions, taking place during learning, in assembly neural networks. The assembly neural network is artificially partitioned into several sub-networks according to the number of classes that the network has to recognize. The features extracted from input data are represented in neural column structures of the sub-networks. Hebbian neural assemblies are formed in the column structure of the sub-networks by weight adaptation. A specific class description is formed in each sub-network of the assembly neural network due to intersections between the neural assemblies. The process of formation of class descriptions in the sub-networks is interpreted as feature generalization. A set of special experiments is performed to study this process, on a task of character recognition using the MNIST database.

show abstract

Object Generation with Neural Networks (When Spurious Memories are Useful)

Cited by 14 publications

References 8 publications

Can the natural system of viruses reconcile the current taxonomy with an alternative classification useful to clinicians?

Can the natural system of viruses reconcile the current taxonomy with an alternative classification useful to clinicians?

The state of play in machine/environment interactions

Generalization of Features in the Assembly Neural Networks

Contact Info

Product

Resources

About