Abstract | ||
---|---|---|
Using the elevator dispatching as an example, we propose a framework for the co-evolution of generic and specific evolutionary neural networks. Neuroevolution is an effective approach to developing many systems, including elevator dispatching, because of the unavailability of desired responses. Furthermore, elevator dispatching is very sensitive to time periods over the course of a day. For such systems, one develop several systems, each of which is specific to a situation or time period. Nevertheless, situations vary day by day. So, we further need to adapt systems to current situations effectively in real time. In this paper, by generating a generic neural network from specific neural networks, we propose the co-neuroevolution of specific and generic neural networks and show its efficiency through simulation. |
Year | DOI | Venue |
---|---|---|
2010 | 10.1109/CEC.2010.5586234 | IEEE Congress on Evolutionary Computation |
Keywords | Field | DocType |
evolutionary neural networks,dispatching,evolutionary computation,neurocontrollers,lifts,adaptive elevator dispatching,control engineering computing,adaptive control,generic neural networks,coneuroevolution,neural nets,adaptive systems,real time,neural network,elevators,artificial neural networks,adaptive system | Adaptive system,Computer science,Evolutionary computation,Elevator,Unavailability,Artificial intelligence,Adaptive control,Artificial neural network,Neuroevolution,Machine learning | Conference |
ISBN | Citations | PageRank |
978-1-4244-6909-3 | 0 | 0.34 |
References | Authors | |
2 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Takahiro Takahashi | 1 | 0 | 0.34 |
Satoshi Matsuda | 2 | 135 | 8.40 |