Relaxation of Coefficient Sensitiveness to Performance for Neural Networks Using Neuron Filter through Total Coloring Problems
スポンサーリンク
概要
- 論文の詳細を見る
In this paper we show that the neuron filter is effective for relaxing the coefficient sensitiveness of the Hopfield neural network for combinatorial optimization problems. Since the parameters in motion equation have a significant influence on the performance of the neural network, many studies have been carried out to support determining the value of the parameters. However, not a few researchers have determined the value of the parameters experimentally yet. We show that the use of the neuron filter is effective for the parameter tuning, particularly for determining their values experimentally through simulations.
- 社団法人電子情報通信学会の論文
- 2001-09-01
著者
-
Higashino Teruo
Department Of Informatics And Mathematical Science Osaka University
-
Higashino Teruo
The Department Of Informatics And Mathematical Science Graduate School Of Engineering Science Osaka
-
Funabiki N
Graduate School Of Natural Science And Technology Okayama University
-
Funabiki Nobuo
The Department Of Informatics And Mathematical Science Graduate School Of Engineering Science Osaka
-
Takenaka Yoichi
The Department Of Informatics And Mathematical Science Graduate School Of Engineering Science Osaka
関連論文
- A WDS Clustering Algorithm for Wireless Mesh Networks
- An Optical-Drop Wavelength Assignment Algorithm for Efficient Wavelength Reuse under Heterogeneous Traffic in WDM Ring Networks(Discrete Mathematics and Its Applications)
- A Minimum Dead Space Algorithm for Generalized Isochronous Channel Reuse Problems in DQDB Networks(Network)
- P2PMM_router : A Two-Stage Heuristic Algorithm to Peer-to-Peer Multicast Routing Problems in Multihome Networks(Discrete Mathematics and Its Applications)
- A Quasi-Solution State Evolution Algorithm for Channel Assignment Problems in Cellular Networks(Special Section on Discrete Mathematics and Its Applications)
- Time-Action Alternating Model for Timed Processes and Its Symbolic Verification of Bisimulation
- A Proposal of Optimal Path Selection Algorithm for Static and Mobile Multicast Routing Problems
- Deriving Concurrent Synchronous EFSMs from Protocol Specifications in LOTOS (Special Section on Selected Papers from the 11th Workshop on Circuits and Systems in Karuizawa)
- A Method to Convert Concurrent EFSMs with Multi-Rendezvous into Synchronous Sequential Circuit(Special Section on Concurrent Systems Technology)
- Execution Time Analysis for Binary Code Executed on a Pipelined Processor Using Parametric Model Checking
- Relaxation of Coefficient Sensitiveness to Performance for Neural Networks Using Neuron Filter through Total Coloring Problems
- A Proposal of Neuron Filter: A Constraint Resolution Scheme of Neural Networks for Combinatorial Optimization Problems
- A Digital Neural Network for Multilayer Channel Routing with Crosstalk Minimization
- A Massive Digital Neural Network for Total Coloring Problems
- A Gradual Neural Network Approach for Time Slot Assignment in TDM Multicast Switching Systems
- Comparisons of Energy-Descent Optimization Algorithms for Maximum Clique Problems (Special Section on Discrete Mathematics and Its Applications)
- A Minimal-State Processing Search Algorithm for Graph Coloring Problems
- A Binary Neural Network Approach for Link Activation Problems in Multihop Radio Networks
- A Neural-Greedy Combination Algorithm for Board-Level Routing in FPGA-Based Logic Emulation Systems(Special Section on Discrete Mathematics and Its Applications)