FaceRec 简单易懂的神经网络面部特征识别例子 [VB源码]

这是我自己花了很多时间写的一个BP神经网络类，里面封装了网络初始化、训练、分类等功能。奉献出来与感兴趣的朋友分享，绝对可靠。-that I spent a lot of time writing in a category BP neural network inside the network initialization package, training, classification capabilities. Dedication out with friends interested in sharing, absolutely reliable.

在matlab6.5下实现的 -achieved in the matlab6.5

he AutoMix package is a C program for Unix-like systems, implementing the automatic reversible jump MCMC sampler of the same name described in Chapters 4, 5, and 6 of David Hastie s Ph.D. thesis-he AutoMix package is a C program for Unix-l ike systems, implementing the automatic reversible jump MC MC sampler of the same name described in Chapter s 4, 5, and 6 of David Hastie"s Ph.D. thesis

一个可自我学习的人工智能小模型，可进行训练，也可进行提问。 希望对大家研究人工智能算法上有帮助！-A small self-learning artificial intelligence model for training, but also ask questions. We hope that research on artificial intelligence algorithms help!

基于canny的灰色预测和模式识别,标定的一种做法是：在一块平板上画若干个点， 同时用这两部相机照相，分别得到这些点在它们像平面上的像点，利用这两组像点的几何关系就可以得到这两部相机的相对位置。然而，无论在物平面或像平面上我们都无法直接得到没有几何尺寸的“点”。实际的做法是在物平面上画若干个圆（称为靶标），它们的圆心就是几何的点了。而它们的像一般会变形，如图1所示，所以必须从靶标上的这些圆的像中把圆心的像精确地找到，标定就可实现。-Based on the gray canny prediction and pattern recognition, calibration approach is to: a plate painted in a number of points, at the same time with these two cameras camera, respectively, as these points in their plane as points, the use of these two groups geometric relationship between the image point can be that the relative positions of two cameras. However, in terms of objects such as plane or plane, we can not directly be no geometry point. Actual practice is complex plane painted in a number of round (known as targets), their geometric center of a circle is the point. And they generally like deformation, as shown in Figure 1, it must target these round, like in the center of a circle of like to find precise, calibration can achieve.

遗传算法求解优化问题，简明教程加示例- The heredity algorithm solution optimization question, the concise course adds the demonstration

用于求解TSP问题的c++程序，已经试运行过-BasicAntColon AlgorithmC++forTSP

网络为３层结构，输入层为６个神经元，隐含层为４个神经元，输出层为１个神经元，当输入层为中心对称样本 1，0，0，0，0，1等时，网络输出为1，否则输出为0。-network of three-layer structure, input layer to six neurons to the hidden layer 4 neurons, the output layer to a neuron, when the input layer to the center symmetry samples such as 1,0,0,0,0,1, the output of a network, the output is 0.

有关TSP遗传算法的的三个pdf论文。不错的参考-The TSP of the three genetic algorithms pdf papers. A good reference

神经网络的运用，比较基础，对初学者比较有用-The use of neural networks, basis of comparison, more useful for beginners