浙江大学编的神经网络入门教程，pdf格式。

基于神经网络的乒乓球机器人乒乓球轨迹预测代码，神经网络的参数可能有待进一步改进-code of PingPong Robot trajectory predicting based on neural network

一个进行数据挖掘的好方法，可以对数据集中的频繁项进行挖掘。但是有一定的缺点，还有需要改进的地方！-Data mining of a better way to focus on data mining of frequent items. Must have the disadvantage, however, also need to improve the place!

一个牛人的神经网络支持向量机程序,对于搞神经网络的朋友强烈推荐.第二部分-A wonderful program of svm code coming from a superman,and strong recommendation to someone who learn about pattern-recognition .The second part.

该进化算法的实现基于VC++6.0的开发平台，主要是用算子改进进化策略算法。-The evolutionary algorithm based on VC++ 6.0 development platform, the main operator is used to improve the evolutionary strategy algorithm.

一个包含丰富内容的流形学习算法工具包，有图形示例文件demo.fig，包括laplacian特征映射算法、流形规则调整、svm分类算法等内容，希望对研究机器学习的朋友有用。-contains a rich content of the manifold learning algorithm tool kit, graphics documents demo.fig example, laplacian features include mapping algorithm manifold rules adjustments svm classification algorithm, want to study machine learning useful friends.

模式识别分类程序，贝叶斯，神经网络分类训练程序，很不错啊-the classification procedures, Bayesian neural network classifier training procedures, and it is very responds :

基于单目手眼相机和激光测距仪，提出了一种尺寸未知的空间矩形平面的位姿测量算法。该算法不需要知道矩形平面 的G 个顶点的物体坐标，只需要知道它们的图像坐标、激光点的图像坐标和激光测距结果，就能够计算出尺寸未知空间矩形平 面在相机坐标系下的位姿，并且计算出矩形平面的尺寸。通过建立单目手眼相机和激光测距仪的数学模型，对该算法进行了验 证。实验结果表明，该算法是有效的，可以应用于机器人对空间物体的跟踪、定位以及抓取。-Monocular-based hand-eye cameras and laser range finder, a size of an unknown two-dimensional rectangular space Pose measurement algorithm. The algorithm does not need to know the G flat rectangular object vertex coordinates, they only need to know the image coordinates, image coordinates of laser points and the laser ranging results, we can calculate the size of the unknown space rectangular plane in the camera coordinate system bit posture, and to calculate the size of rectangular plane. Through the establishment of monocular hand-eye cameras and laser range finder of the mathematical model, the algorithm is verified. Experimental results show that the algorithm is effective, can be applied to the robot on the space object tracking, positioning, and crawling.

是实现关系型贝叶斯网络一中机器学习算法，从国外相关研究人员的主页上下载收集。

模式识别与人工智能，圆检测HOUGH变换的改进算法-Pattern Recognition and Artificial Intelligence, HOUGH transform circle detection algorithm to improve

遗传算法优化神经网络结构和链接权，采用浮点数编码-Genetic algorithm neural network structure and link the right to use floating-point encoding