Title :
Sparsely Connected Associative Memory with Adaptive Topology through Annealed Dilution
Author :
Jing Yang ; Bin Kong
Author_Institution :
Inst. of Intell. Machines, Hefei, China
Abstract :
A novel sparsely connected associative memory model with adaptive topology through annealed dilution is proposed in this paper. Aimed at overcoming the disadvantage of quenched dilution as random synapses disconnection of the existing methods, this new model takes the ideology of annealed dilution into account, and investigates the optimal synaptic dilution strategy under the constraints of limited amount of neurons and connections. Based on explicit theoretical analysis, this new model breaks the traditional manner of quenched dilution but instead constructs a learning task-dependent network topology through annealed way which is much closer to biological genuine system as possessing flexible adaptive structure and can also achieve better performance than the existing counterparts of the same class. the rationality and validity of the proposed model is validated from great number of experiments.
Keywords :
content-addressable storage; learning (artificial intelligence); neural nets; topology; adaptive topology; annealed dilution; biological genuine system; flexible adaptive structure; learning task-dependent network topology; neurons; optimal synaptic dilution strategy; quenched dilution; random synapses disconnection; sparsely connected associative memory model; Adaptation models; Annealing; Associative memory; Network topology; Neurons; Noise; Topology; adaptive topology; annealed dilution; associative memory; sparsely connected;
Conference_Titel :
Computational Intelligence and Design (ISCID), 2012 Fifth International Symposium on
Conference_Location :
Hangzhou
Print_ISBN :
978-1-4673-2646-9
DOI :
10.1109/ISCID.2012.50