2012
DOI: 10.1109/tnnls.2011.2178311
|View full text |Cite
|
Sign up to set email alerts
|

Tackling Learning Intractability Through Topological Organization and Regulation of Cortical Networks

Abstract: A key challenge in evolving control systems for robots using neural networks is training tractability. Evolving monolithic fixed topology neural networks is shown to be intractable with limited supervision in high dimensional search spaces. Common strategies to overcome this limitation are to provide more supervision by encouraging particular solution strategies, manually decomposing the task and segmenting the search space and network. These strategies require a supervisor with domain knowledge and may not be… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1

Citation Types

1
39
0

Year Published

2014
2014
2023
2023

Publication Types

Select...
6
3
1

Relationship

0
10

Authors

Journals

citations
Cited by 18 publications
(40 citation statements)
references
References 27 publications
1
39
0
Order By: Relevance
“…Among them, the modeling of neuronal networks of a brain can be viewed as a typical application of complex networks [2], [3]. In [4]- [6], the theoretical modeling, tackling learning of neuronal networks and the applications of neuronal networks to image processing have been investigated, respectively. Modern brain mapping approaches such as diffusion MRI, functional MRI, EEG, and MEG have constantly produced large datasets of anatomical and functional connection patterns.…”
Section: Introductionmentioning
confidence: 99%
“…Among them, the modeling of neuronal networks of a brain can be viewed as a typical application of complex networks [2], [3]. In [4]- [6], the theoretical modeling, tackling learning of neuronal networks and the applications of neuronal networks to image processing have been investigated, respectively. Modern brain mapping approaches such as diffusion MRI, functional MRI, EEG, and MEG have constantly produced large datasets of anatomical and functional connection patterns.…”
Section: Introductionmentioning
confidence: 99%
“…SphereX has the same goals as the Microbots, but with the goal of launching fewer robots, that are better equipped with sciencegrade instruments. Moreover, the path planning approach addressed in this paper is motivated by multiple hopping robots navigating a maze [38], where one robot hops at a time within a local area of safety, collects more information and plans for their next hop. Our past work has shown has shown the feasibility of multiple small robots working together as a network [25], [26], [27], [28].…”
Section: Related Work and Motivationmentioning
confidence: 99%
“…AMIGO is advancement over our nano-lander concept, as it contains a multi-functional inflatable [29][30][31][32] for use in mobility, communications and tracking and has a propulsion system finely tuned to the asteroid low-gravity environment [26]. AIMGO by hopping and rolling will be able to navigate the maze-like surface pathways using autonomous control [36].…”
Section: Introductionmentioning
confidence: 99%