感知器的贝叶斯在线算法性能
Performance of the Bayesian online algorithm for the perceptron.
作者信息
de Oliveira Evaldo Araújo, Alamino Roberto Castro
出版信息
IEEE Trans Neural Netw. 2007 May;18(3):902-5. doi: 10.1109/TNN.2007.891189.
PMID:17526354
Abstract
In this letter, we derive continuum equations for the generalization error of the Bayesian online algorithm (BOnA) for the one-layer perceptron with a spherical covariance matrix using the Rosenblatt potential and show, by numerical calculations, that the asymptotic performance of the algorithm is the same as the one for the optimal algorithm found by means of variational methods with the added advantage that the BOnA does not use any inaccessible information during learning.
摘要
在这封信中，我们使用罗森布拉特势推导了具有球形协方差矩阵的单层感知器的贝叶斯在线算法（BOnA）泛化误差的连续方程，并通过数值计算表明，该算法的渐近性能与通过变分方法找到的最优算法相同，此外还有一个额外的优点，即BOnA在学习过程中不使用任何不可获取的信息。
相似文献
1
Performance of the Bayesian online algorithm for the perceptron.感知器的贝叶斯在线算法性能
IEEE Trans Neural Netw. 2007 May;18(3):902-5. doi: 10.1109/TNN.2007.891189.
2
The Rosenblatt Bayesian algorithm learning in a nonstationary environment.罗森布拉特贝叶斯算法在非平稳环境中的学习。
IEEE Trans Neural Netw. 2007 Mar;18(2):584-8. doi: 10.1109/TNN.2006.889943.
3
Variational Bayesian approach to canonical correlation analysis.变分贝叶斯方法用于典型相关分析。
IEEE Trans Neural Netw. 2007 May;18(3):905-10. doi: 10.1109/TNN.2007.891186.
4
Unsupervised learning of gaussian mixtures based on variational component splitting.基于变分分量分裂的高斯混合无监督学习。
IEEE Trans Neural Netw. 2007 May;18(3):745-55. doi: 10.1109/TNN.2006.891114.
5
An empirical evaluation of the fuzzy kernel perceptron.模糊核感知器的实证评估。
IEEE Trans Neural Netw. 2007 May;18(3):935-7. doi: 10.1109/TNN.2007.891624.
6
Global asymptotic stability of delayed cellular neural networks.时滞细胞神经网络的全局渐近稳定性
IEEE Trans Neural Netw. 2007 May;18(3):947-50. doi: 10.1109/tnn.2007.891628.
7
A geometrical method to improve performance of the support vector machine.一种提高支持向量机性能的几何方法。
IEEE Trans Neural Netw. 2007 May;18(3):942-7. doi: 10.1109/TNN.2007.891625.
8
New dynamical optimal learning for linear multilayer FNN.线性多层前馈神经网络的新型动态最优学习
IEEE Trans Neural Netw. 2004 Nov;15(6):1562-8. doi: 10.1109/TNN.2004.830801.
9
A convex approach to validation-based learning of the regularization constant.一种基于验证的正则化常数学习的凸方法。
IEEE Trans Neural Netw. 2007 May;18(3):917-20. doi: 10.1109/TNN.2007.891187.
10
Knowledge extraction from neural networks using the all-permutations fuzzy rule base: the LED display recognition problem.使用全排列模糊规则库从神经网络中提取知识：LED显示屏识别问题。
IEEE Trans Neural Netw. 2007 May;18(3):925-31. doi: 10.1109/TNN.2007.891686.
Zotero 插件