| 基于Bayesian的期望最大化方法——BEM算法 |
| |
| 引用本文： | 温津伟,罗四维,赵嘉莉,韩臻.基于Bayesian的期望最大化方法——BEM算法[J].计算机研究与发展,2001,38(7):821-825. |
| |
| 作者姓名： | 温津伟 罗四维 赵嘉莉 韩臻 |
| |
| 作者单位： | 北方交通大学计算机科学技术学院 |
| |
| 基金项目： | 国家自然科学基金项目资助 (6 99730 0 2 ) |
| |
| 摘 要： | 通过对标准EM算法收敛于局部极值的原因进行分析，提出了基于Bayesian方法的神经网络新学习算法－－BEM算法。该算法解决了标准EM算法的上述缺陷，同时还可防止标准EM算法Overfitting情况的出现，并可防止标准EM算法有时只响应第一模式而失去泛化能力情况的出现。实验结果表明了该算正确性和有效性。该算法对研究和发展标准EM学习算法理论具有一定的学习意义。
|
| 关 键 词： | 随机神经网络 EM算法 Bayesian方法 Wishart－Gaussian分布 |
THE BEM ALGORITHM: AN EM METHOD BASED ON BAYESIAN |
| |
| WEN Jin Wei,LUO Si Wei,ZHAO Jia Li,and HAN Zhen.THE BEM ALGORITHM: AN EM METHOD BASED ON BAYESIAN[J].Journal of Computer Research and Development,2001,38(7):821-825. |
| |
| Authors: | WEN Jin Wei LUO Si Wei ZHAO Jia Li and HAN Zhen |
| |
| Abstract: | A new neural network learning algorithm based on Bayesian method (BEM algorithm) is presented. One disadvantage of the standard EM algorithm is its convergence to local minimum. The BEM algorithm overcomes the disadvantage by analysing the reasons. Furthermore, BEM avoids the EM's overfitting problem as well as its disability of generalization due to only responding to single patterm. Experimental result shows the BEM's correctness and validity. The BEM has academic values of contributing to the research and development of EM algorithm. |
| |
| Keywords: | random neural networks EM algorithm Bayesian method Wishart Gaussian distribution |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
