A multilevel neuromolecular architecture that uses the extradimensional bypass principle to facilitate evolutionary learning
References (35)
Curr. Mod. Biol.
(1972)J. Theor. Biol.
(1974)BioSyst.
(1976)J. Theor. Biol.
(1977)Bull. Math. Biol.
(1979)BioSyst.
(1979)BioSyst.
(1982)BioSyst.
(1982)BioSyst.
(1989)Molecular computing
BioSyst.
BioSyst.
Physica D
Biosyst.
BioSyst.
Brain Res.
Physica D
Cited by (33)
Modeling the evolution of molecular systems from a mechanistic perspective
2014, Trends in Plant ScienceCitation Excerpt :Therefore, crossing fitness valleys through point mutation trajectories is often not feasible, and special evolutionary mechanisms need to be invoked to create passable ridges on the fitness landscape. In particular, the navigability of sequence space may be facilitated by introducing extra genotypic dimensions, enabling the circumvention of reciprocal sign-epistatic effects by rerouting the evolutionary trajectory (i.e., by making compensatory changes at other loci first, a mechanism that has been termed an ‘extradimensional bypass’ [90–92]). Although increased navigability has been proposed to be a general feature of high-dimensional fitness landscapes [93,94], it has been linked to gene duplication in particular, because a duplicated gene copy may buffer deleterious mutations in the other copy under some conditions [91,92,95] (Figure 3).
Design and implementation of an artificial neuromolecular chip and its applications to pattern classification problems
2009, NeurocomputingCitation Excerpt :The transduction mechanism of cytoskeletal neurons is motivated by some physiological evidence that the intraneuronal dynamics of a neuron control its firing behavior [35–37]. High dimensionality and interactional complexity can enhance the evolvability of the cytoskeletal neurons [38,39]. The reference neuron scheme is basically a Hebbian model, in which the connection between two neurons is strengthened when they are active simultaneously.
Neuromolecularware and its application to pattern recognition
2009, Expert Systems with ApplicationsData differentiation and parameter analysis on the weight changes of premature babies with an artificial neuromolecular system
2008, Expert Systems with ApplicationsExtradimensional bypass
2002, BioSystemsTechniques for enhancing neuronal evolvability
2002, NeurocomputingCitation Excerpt :We should also note that the alternating mode of variation–selection (i.e., varying distinct parameters in different phases) appears to allow for much faster evolution than a simultaneous mode in which all parameters are subject to variation at once. This result was also obtained with the artificial neuromolecular architecture previously noted [6–9] and with other models. Whether biological systems have some kind of hierarchical genetic structure variation and selection to focus on different phenotypic features during different phases of natural evolution is an open question.