How Hierarchical Control Self-organizes in Artificial Adaptive Systems Export

@article{citeulike:927180, abstract = {Diverse, complex, and adaptive animal behaviors are achieved by organizing hierarchically structured controllers in motor systems. The levels of control progress from simple spinal reflexes and central pattern generators through to executive cognitive control in the frontal cortex. Various types of hierarchical control structures have been introduced and shown to be effective in past artificial agent models, but few studies have shown how such structures can self-organize. This study describes how such hierarchical control may evolve in a simple recurrent neural network model implemented in a mobile robot. Topological constraints on information flow are found to improve system performance by decreasing interference between different parts of the network. One part becomes responsible for generating lower behavior primitives while another part evolves top-down sequencing of the primitives for achieving global goals. Fast and slow neuronal response dynamics are automatically generated in specific neurons of the lower and the higher levels, respectively. A hierarchical neural network is shown to outperform a comparable single-level network in controlling a mobile robot. 10.1177/105971230501300303}, author = {Paine, Rainer W. and Tani, Jun}, citeulike-article-id = {927180}, citeulike-linkout-0 = {http://dx.doi.org/10.1177/105971230501300303}, citeulike-linkout-1 = {http://adb.sagepub.com/content/13/3/211.abstract}, citeulike-linkout-2 = {http://adb.sagepub.com/content/13/3/211.full.pdf}, citeulike-linkout-3 = {http://adb.sagepub.com/cgi/content/abstract/13/3/211}, day = {1}, doi = {10.1177/105971230501300303}, journal = {Adaptive Behavior}, keywords = {evolutionary, hierarchical, neural\_network}, month = {September}, number = {3}, pages = {211--225}, posted-at = {2007-10-02 11:17:32}, priority = {3}, title = {How Hierarchical Control Self-organizes in Artificial Adaptive Systems}, url = {http://dx.doi.org/10.1177/105971230501300303}, volume = {13}, year = {2005} }

TY - JOUR ID - citeulike:927180 L3 - citeulike-article-id:927180 N2 - Diverse, complex, and adaptive animal behaviors are achieved by organizing hierarchically structured controllers in motor systems. The levels of control progress from simple spinal reflexes and central pattern generators through to executive cognitive control in the frontal cortex. Various types of hierarchical control structures have been introduced and shown to be effective in past artificial agent models, but few studies have shown how such structures can self-organize. This study describes how such hierarchical control may evolve in a simple recurrent neural network model implemented in a mobile robot. Topological constraints on information flow are found to improve system performance by decreasing interference between different parts of the network. One part becomes responsible for generating lower behavior primitives while another part evolves top-down sequencing of the primitives for achieving global goals. Fast and slow neuronal response dynamics are automatically generated in specific neurons of the lower and the higher levels, respectively. A hierarchical neural network is shown to outperform a comparable single-level network in controlling a mobile robot. 10.1177/105971230501300303 IS - 3 JF - Adaptive Behavior EP - 225 TI - How Hierarchical Control Self-organizes in Artificial Adaptive Systems VL - 13 SP - 211 KW - evolutionary KW - hierarchical KW - neural_network AU - Paine, Rainer AU - Tani, Jun PY - 2005/09/01/ UR - http://dx.doi.org/10.1177/105971230501300303 ER -