| BP神经网络的优化算法研究 |
| |
| 引用本文： | 张山,何建农.BP神经网络的优化算法研究[J].计算机与现代化,2009(1). |
| |
| 作者姓名： | 张山 何建农 |
| |
| 作者单位： | 福州大学数学与计算机科学学院,福建,福州,350002 |
| |
| 摘 要： | BP学习算法通常具有收敛速度慢,易陷入局部极小值等缺点;遗传算法是全局优化算法,具有较强的全局搜索性能,但它在实际应用中容易产生早熟收敛的问题,且在进化后期搜索效率较低;模拟退火算法具有摆脱局部最优点的能力,能抑制遗传算法的早熟现象.因此,本文在BP算法结合遗传算法的同时,再加入模拟退火算法,可以有效地缓解遗传算法的选择压力.
|
| 关 键 词： | 遗传算法 模拟退火算法 神经网络 |
Research on Optimized Algorithm for BP Neural Networks |
| |
| ZHANG Shan,HE Jian-nong.Research on Optimized Algorithm for BP Neural Networks[J].Computer and Modernization,2009(1). |
| |
| Authors: | ZHANG Shan HE Jian-nong |
| |
| Affiliation: | College of Mathematics and Computer Science;Fuzhou University;Fuzhou 350002;China |
| |
| Abstract: | BP learning algorithm converges slow and the solution got is usually local optimal solution.Genetic Algorithm is global optimization algorithm and has strong global search ability.But it can easily cause premature convergence problems in practical application and the efficient of search in later evolution is low.Simulted Annealing Algorithm has the advantages of avoid getting the local optimal solution and premature convergence problem.So BP algorithm combines Genetic Algorithm and Simulated Annealing Algor... |
| |
| Keywords: | genetic algorithm simulated annealing algorithm BP neural networks |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
