OPTIMAL GENERALIZATION IN PERCEPTRONS
OPTIMAL GENERALIZATION IN PERCEPTRONS
复制标题
DOI：
10.1088/0305-4470/25/23/020
复制
发表时间：
1992-12-07
期刊：
影响因子：
--
通讯作者：
CATICHA, N
中科院分区：
文献类型：
--
作者：
KINOUCHI, O;CATICHA, N
提出了一种新的单层感知器学习算法。它的目标是最大化每个示例的泛化增益。针对每个示例单独呈现的情况获得了分析结果。赫布项的权重是教师感知器中示例的预期稳定性的函数。这样就得到了泛化能力的上限。该方案可以迭代，数值模拟的结果表明，它在误差范围内收敛于贝叶斯算法理论上的最优泛化能力。通过实例选择的学习策略，得到了具有最大化泛化能力的算法的分析和数值结果，并证明了如预期的那样，正交选择是最优的。对于所选示例的单一呈现，获得了泛化误差的指数衰减。
A new learning algorithm for the one-layer perceptron is presented. It aims to maximize the generalization gain per example. Analytical results are obtained for the case of single presentation of each example. The weight attached to a Hebbian term is a function of the expected stability of the example in the teacher perceptron. This leads to the obtention of upper bounds for the generalization ability.This scheme can be iterated and the results of numerical simulations show that it converges, within errors, to the theoretical optimal generalization ability of the Bayes algorithm.Analytical and numerical results for an algorithm with maximized generalization in the learning strategy with selection of examples are obtained and it is proved that, as expected, orthogonal selection is optimal. Exponential decay of the generalization error is obtained for the single presentation of selected examples.