基于遗传算法的模拟退火

数据挖掘中的关联规则发现算法-Data Mining Association Rules algorithm

以黎曼几何为理论依据，基于S.Amari的修正核函数思想提出了两种新的保角变换，用其对核函数进行数据依赖性改进，进一步提高支持向量机分类器泛化能力。以人工非线性分类问题 为对象进行研究，仿真实验结果表明采用新保角映射可以快速显著地改善分类器泛化性能，而且能大幅度地减少支持向量的数目。-Two novel conformal transformations were proposed based on the Riemannian geometry theory and S.Amari’sidea.And the kernel function was modified by the transformation in a data-dependent way.Our experimental results for theartificial nonlinear data set show that the generalization performance of support vector machines classifier is improvedremarkably and the number of support vectors is decreased greatly.

模型自适应模糊控制源代码，基于matlab编程，并给出了仿真实例，具体说明见程序。-self-adapted mode of blured control program,based on matlab programming,and also have a emluator,and can concretely see program s explain

利用双层感知器神经网络处理线性不可分的分类问题-use double perceptron neural networks to handle non-linear classification

BP神经网络解决异或问题-BP neural network or problems to solve differences

神经网络控制作为一种全新的智能控制方法,是解决非线性系统预测和控制问题的一种重要手段,受到了许多专家学者的广泛关注,是目前国内外研究的热点之一。本文着重研究前向神经网络的学习算法,简单探讨了BP算法在逼近非线性系统中各个因素对系统的影响。 -Neural Network Control as a new intelligent control method is to solve the nonlinear system to predict and control an important means by many experts and scholars in a wide range of concerns, is one of the hot research. This article focuses on feedforward neural network learning algorithm, a simple BP algorithm of approximation of nonlinear systems in the various factors on the system.

基本遗传算法的源代码，比较各代种群的进化。-The basic genetic algorithm source code, comparing the generation of species evolution.

用BCB编写的Kohonen程序做优雅的网格分布，将任意维的输入数据，简化为二维x，y平面，竞争学习最佳实践。

every function has its own help which can be obtained by typing "help name" where "name" is the function, e.g. "help fcm"-every function has its own help which can be obtained by typing "help name" where "name" is t he function, eg "help FCM"

这是一个使用VC++编写的“基于神经网络方法的专家系统源代码”，很有参考价值，请检查-It is a preparation of the VC, "based on neural network expert system source code," was useful, please check