一个外国人写的字符识别训练源码，VC++源码，有源码、说明及演示可执行程序，对学习和研究模式识别很有帮助。...

遗传算法,实现的是一个猜答案的过程， 这种算法比较经典，可以借鉴-genetic algorithm to achieve guess the answer is a process, this method is classic, we can draw on

神经网络训练算法,很快 很简单 很适用 能快速用到实验环境中去-Neural network training algorithm, is applicable to very quickly is very simple to quickly use the experimental environment

这是一个指纹识别的源代码,指纹识别是比较成熟的模式识别!`-This is a fingerprint identification of the source code, fingerprint identification is a relatively mature pattern recognition! `

Fisher线性判别分类器应用于IRIS数据集的例子-An example of Fisher linear discriminant Classifier applied on IRIS Data Set

有关摄像头定标的C++代码，标定摄像头的12个参数-Related to camera calibration C++ Code, camera calibration parameters 12

围绕粒子群的当前质心对粒子群重新初始化.这样，每个粒子在随后的迭代中将在新的位置带着粒子在上次搜索中获得的“运动惯性”（wvi）向Pi，Pg的方向前进，从而可以在粒子群的运动过程中获得新的位置，增加求得更优解的机会.随着迭代的继续，经过变异的粒子群又将趋向于同一点，当粒子群收敛到一定程度时又进行下一次变异，如此反复，直到迭代结束.-particle swarm around the center of mass of the current PSO reinitialization. Thus, Each particle in the next iteration will be in the new location with particles in the last search was the "inertia" (wvi ) Pi, Pg orientation, and thus can PSO course of the campaign was a new position, increase seek better solutions opportunities. With the continued iteration, after variation of PSO will tend to the same point. When PSO converge to a certain extent when the next variation, so repeatedly, until the end of iteration.

这是一个神经网络的开发工具-This is a neural network development tools

提出了基于最小化估计的泛化误差界非优化估计方法。目前大多数的核参数选择方法都是通过极小化了来得到最优参数值，但是求解优化问题的计算代价相当的大，并且不能很好地体现数据的分布特征。本文采用非优化技术，通过极小化泛化误差来优化核及相关参数，由于直接计算最小半径和最大间隔，避免了对优化问题的直接求解，因此可以很好地降低计算代价。并且该方法直接从样本出发，可以很好地体现数据的分布特征，不管数据分布是否均匀都可以适用。给出了基于凸包估计的SVM核选择的模型及实现算法。 -Proposed based on minimizing the estimated generalization error bound of the non-optimal estimation method. Most of the nuclear parameter selection methods are to get the best by minimizing the parameter values, but the computational cost for solving optimization problems is quite large, and can not properly reflect the distribution characteristics of the data. In this paper, the non-optimization technology, by minimizing the generalization error to optimize the nuclear and related parameters, due to the direct calculation of the minimum radius and maximum interval, avoiding the direct solution of the optimization problem, it can very well reduce the computation cost. And that the method directly from the sample one can well reflect the distribution chara

这是一个模式识别中的留一法算法的一个简单仿真分类实例，其中female.txt和male.txt是训练样本，test.txt是测试样本，分类效果非常好，对于模式学习的初学者将会有很大帮助。-This is a pattern recognition algorithm for leave-one simulation of a simple classification example, is one of female.txt and male.txt training samples, test.txt is the measurement, classification effect is very good for beginners will be learning model there is a great help.

目前开发人工只能的朋友一定会遇到bp算法中八进制的问题，这段程序很好的解决了这个问题。我目前在作模糊控制，神经网络，遗传算法的课题，收集了很多相关程序，望同方向者探讨一二。-only friends will find bp algorithm octal problem, this process satisfactorily resolved this problem. Now I make fuzzy control, neural networks, genetic algorithms topic, a lot of collection procedures, and hopes to explore the same direction on January 2.