资源列表
ica453
- Demonstration and test of the kernel-ica package
LVQ
- LVQ(学习矢量量化)算法：它是Kohonen的有监督学习的扩展形式。融合了自组织和有导师监督的技术，学习方法是竞争的，但产生方式是有教师监督的，也就是说，竞争学习是在由训练输入指定的各类中局部进行。-LVQ (LVQ) algorithm : it is Kohonen of supervised learning the expansion of the form. The convergence of self-organization and supervision of the ins
RLCX
- Kohonen的SOFM(自组织特征映射)：这种算法部分收到生物特征影响，在网络输出层内按几何中心或特征进行聚类。-Kohonen the SOFM (self-organizing feature map) : This algorithm received some biological features, the output layer to the network within the geometric center or cluster features.
elman6
- 基于ELMAN神经网络对阶跃响应动态进行补偿也可以举一反三的运用于ELMAN网络的其他应用-Elman neural network based on the step response of the dynamic compensation can be applied to an example of the Elman network other applications
基于AI的问题求解
- 提供了深度优先搜索,广度优先搜索,添加启发信息的最小代价搜索,查找多个解,查找最优解的范例程序.-a depth-first search, breadth - first search, add the information inspired the minimum price for a search of solutions to search, find out the optimal solution sample programs.
人脸识别的核心源代码
- 人脸识别的核心源代码，PDF格式文件，采用特征脸方法。可供有兴趣的人阅读。-the core source code, documents in PDF format, the method adopted Eigenface. For those who are interested in reading.
第四章 控制系统的分析方法
- 用Canny算子检测图像的边缘 P0404：图像的阈值分割 P0405：用水线阈值法分割图像 P0406：对矩阵进行四叉树分解 P0407：将图像分为文字和非文字的两个类别 P0408：形态学梯度检测二值图像的边缘 P0409：形态学实例——从PCB图像中删除所有电流线，仅保留芯片对象-with Canny operator to detect the edges in the image P0404 : image thresholding segmentation P0405 : water
speechrecognition
- 语音信号处理工具包,里面提供了很多有用的操作。这是最新版本-voice signal processing tool kit containing a lot of useful operation. This is the latest version
hmm123
- hmm算法 包括前后向算法、viterbi算法、bw算法等另外包括内存申请等源代码-algorithm to include before and after algorithm, Viterbi algorithm, bw algorithm other memory applications including source code
function_opt_demo
- 使用matlab应用遗传算法进行函数优化-Using genetic algorithm in matlab to optimize function
Matrix2003
- 矩阵计算库，经改编过的，可以扩充，可读性很好。-matrix calculation library, adapted too, can be expanded, good readability.
Gen2005
- 遗传算法的一个库，算法设置比较灵活，需要懂算法的人来下载-a genetic algorithm library, set up a more flexible algorithm, the algorithm needs to understand people to download