PSO算法求解带约束优化问题，包含所有源程序和运算图片。写论文是编的。-PSO algorithm for solving constrained optimization problems, including all source code and the computing picture. Writing papers is made of.

基于倒谱特征的带噪语音端点检测.rar格式为ｖｉｐ-based Cepstral Noisy voice endpoint detection. Rar format vip

人工智能中重要的一个问题, 用广度优先搜索的方法解决-AI an important issue, breadth- first search of solutions

本文采用红外光电传感器技术, 建立了一个基于神经网络的客流计数系统, 与传 统的仅利用光电传感器计数方法相比, 该方法提高了实时客流计数准确率。为了更好的分 类识别计数, 针对采集的连续的时空序列客流数据特点, 改进了已有的分割方法和针对客 流实际情况的特征提取方法。最后经过稀疏和密集客流模拟仿真测试, 结果表明, 本文所 提出的模型能够识别多人并排的情况, 误差率较低, 具有理论和实际意义。-RBF

贝叶斯算法是基于贝叶斯定理 P(H|X) = P(X|H)P(H) / P(X).。对于多属性的数据集，计算 P(X|Ci) 的开销非常大，为减低计算复杂度，我们做条件独立的假设，即给定元组的类标号，假定属性值有条件地相互独立，即在属性间不存在依赖关系。此程序仅为算法的一个实现，根据训练数据训练分类器-Bayesian algorithm is based on the Bayes theorem P (H | X) = P (X | H) P (H)/P (X).. For multi-attribute data sets, computing P (X | Ci) of the overhead is very large, in order to reduce the computational complexity, we do conditional independence assumption that a given tuple class label, it is assumed that property values conditionally independent of each other, that does not exist in the inter-attribute dependencies. This procedure is only an implementation of algorithm, according to training data classifier training

汉字识别系统，非常简单，网络上看到的 一起学学啊，对模式的一个简单匹配.-Chinese character recognition system is very simple, see the Web together to learn ah, the model of a simple matching.

使用线性分类器进行分类，采用感知器算法中的“奖惩算法”，各提取3类中的前25个样本共75个作为学习样本。-The use of a linear classifier for classification, perceptron algorithm using the reward and punishment algorithm , the extraction of 3 categories of the top 25 a total of 75 samples as the learning samples.

Fuzzy Rule Base Implementation

基于BP算法的贝叶斯网络参数学习， 实现了图形话的节点表示和样本生成， 图形话的结果分析-BP algorithm based on Bayesian network parameter learning, the achievement of the graphics, then the node express and sample generation, if the results of graphical analysis

VC写的harris角点检测程序。又名plessey，最经典也是目前效果最好的算法，比susan抗噪要好（实际的讲）-A program of Harris corner points test writen by VC.We can call is as plessey,it is the best algorithm.It better than susan from real useful point.