| GA-SAPSO神经网络模型的构建 |
| |
| 引用本文： | 周建新,付传秀.GA-SAPSO神经网络模型的构建[J].佳木斯大学学报,2011,29(1):32-35. |
| |
| 作者姓名： | 周建新 付传秀 |
| |
| 作者单位： | 皖西学院文科综合实训中心;皖西学院应用数学学院; |
| |
| 摘 要： | BP神经网络存在寻优参数多、收敛速度慢、易陷入局部极小的固有缺陷,为改进其网络性能,本文利用遗传-模拟退火粒子群算法(GA-SAPSO)对BP神经网络的初始权值及神经元阀值进行优化处理,并重新构建网络模型.实例仿真结果表明:所构建模型降低了BP网络结构的复杂性,避免了网络参数选取的盲目性,提高了网络的计算精度.
|
| 关 键 词： | 遗传算法 模拟退火粒子群算法 BP网络 优化 |
The Model Construction of GA-SAPSO Neural Network |
| |
| ZHOU Jian-xin,FU Chuan-xiu.The Model Construction of GA-SAPSO Neural Network[J].Journal of Jiamusi University(Natural Science Edition),2011,29(1):32-35. |
| |
| Authors: | ZHOU Jian-xin FU Chuan-xiu |
| |
| Affiliation: | ZHOU Jian-xin1,FU Chuan-xiu2(1.Modern Education Center,West Anhui University,Luan 237012,China,2.College of Applied Mathematics,China) |
| |
| Abstract: | According to the inherent defects of slow convergence speed and easily falling into the local minimum on BP neural network,this paper constructed a model using GA-SAPSO to optimize the network weights and neuron thresholds,then integrated them as the initial weights and thresholds to rebuild BP network.The simulation results show that this model can reduce the structure complexity,avoid the blind parameters choice and improve calculation accuracy of the BP neural network. |
| |
| Keywords: | genetic algorithm simulated annealing-particle swarm optimization BP network optimization |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
