Study of a Neural Network with a Meta-layer

Loocke, Philip R. Van

doi:10.1080/09540099108946593

Cited by 2 publications

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Categorization and the resulting hierarchical structure of knowledge is dealt with by others in this volume and elsewhere [e.g. Van Loocke 1991]. The topic is not addressed in great depth here, tough it is of relevance to point out that as a consequence of chunking/categorization, the complexity of what can be held in the focus (and thus of what constitutes a meme, and thus a feature) will differ amongst individuals, and within an individual over time.…”

Section: Meme As Pattern Of Information Encoded In the Focusmentioning

confidence: 99%

A Day in the Life of a Meme.

Gabora¹

1996

Philosophica

View full text Add to dashboard Cite

Like the infonnation patterns that evolve through. biological processes, mental representations or memes evolve through adaptive exploration and transfonnation of an infonnation space through variation, selection, and transmission. However since memes do not contain instructions for their replication our brains do it for them, strategically, guided by a fitness landscape that reflects both internal drives and a worldview that fonns through meme assimilation. This paper presents a tentative model for how an individual becomes a memeevolving agent via the emergence of an autocatalytic network of sparse, distributed memories, and discusses implications for complex creative thought processes and why they are unique to humans. A hypothetical scenario for the evolutionary dynamics of a given meme in a society of interacting individuals is presented.

show abstract

Section: Meme As Pattern Of Information Encoded In the Focusmentioning

confidence: 99%

A Day in the Life of a Meme.

Gabora¹

1996

Philosophica

View full text Add to dashboard Cite

show abstract

The Connectionist Model QNET and its Combination with Genetic Algorithms.

Loocke¹

1996

Philosophica

View full text Add to dashboard Cite

This paper starts with an exposition of a new connectionist model. The model genemlizes more classical fully connected symrr:tetric networks. One of its main characteristics is that the state of a unit is characterized by a vector of amplitudes rather than by a single scalar . . activation value. The rules that express the evolution of the amplitudes of a particular unit are expressed. in terms of the amplitudes of the other units. The chance to observe a particular unit as being active in a particular frequency, however, is proportional with the square of the corresponding amplitude. Because of this as well as some other parallels with quantum theory, the model has been called QNET. QNET operates in different frequencies at once, and the operations in different frequencies interact with each other in a way that is desimble from a cognitive point of view. For instance, when confronted with a problem, QNET can find different solutions in differen~ frequencies. When different solutions are found, this often indicates that more classical networks fail to find a solution at all, since they then converge to a spurious mixture of solutions. We go on to consider how this type of network can be combined with genetic algorithms. We point out that a combination• of both methods leads to a technique that integrates the benefits of genetic algorithms with the ones of neural networks. .

show abstract

Study of a Neural Network with a Meta-layer

Cited by 2 publications

References 17 publications

A Day in the Life of a Meme.

A Day in the Life of a Meme.

The Connectionist Model QNET and its Combination with Genetic Algorithms.

Contact Info

Product

Resources

About