KNN和BP的算法实现。输入一系列的样本值，进行训练和分类。源码是实现了两个分类器，已经调试通过

matlab编写的BP神经网络算法，的一个应用小程序-matlab prepared by BP neural network algorithm, a small process applications

Robocup机器人小程序,实现了机器人的射门,但机器人之间的配合没有实现-Robocup robot small programs to achieve the goal of a robot, but the robot between the not achieved

基于多目标优化的免疫遗传算法 在Matlab环境中的实现,期刊论文。-Multi-objective optimization based on immune genetic algorithm in Matlab environment, the realization of journal articles.

一个求解TSP问题的例子,用delphi编写.有很好的参考价值 -a TSP example, prepared with delphi. A very good reference value

禁忌搜索法对初始解、邻域个数及禁忌列表的大小等参数有比较严格的要求, 这些参数直接影响着算法的优化能力。文章提出了一种改进的禁忌搜索法, 它用有效空间来压缩搜索范围, 这样可以提高搜索效率和全局搜索能力。用短期和长期禁忌列表存储器来保证算法能搜索到全和分局空间的每一点, 并且不重复搜索。经过验算析, 证明它是一种较好的全局启发式搜索法.-Tabu search method, the initial solution, neighborhood and tabu list size of the number of parameters such as have more stringent requirements, these parameters directly affect the ability to optimize the algorithm. This paper presents an improved tabu search method, it is used effectively to compress the search space, so that can improve search efficiency and global search capabilities. Short-term and long-term memory tabu list to ensure the algorithm can search the whole and the sub-space, each point, and do not repeat the search. After checking analysis to prove that it is a good global heuristic search method.

神经网络相关算法，使用java语言进行描述的-Neural network-related algorithms, using the java language to describe the

c语言编写的蚁群算法源程序，应用于TSP-c language source code of ant colony algorithm applied to TSP

图像模式识别的各种算法及其编程实现,包括模板匹配分类器、几何分类器和神经网络分类器等，可用于手写数字的识别。-A variety of image pattern recognition algorithm and its programming, including template matching classifier, geometric classifier and neural network classifiers, etc., can be used for handwritten numeral recognition.

基于自组织特征映射网络的聚类分析，是在神经网络基础上发展起来的一种新的非监督聚类方法，分析了基于自 组织特征映射网络聚类的学习过程，分析了权系数自组织过程中邻域函数和学习步长的一般取值问题，给出了基于自组织 特征映射网络聚类实现的具体算法，并通过实际示例测试，证实了算法的正确性。 -Based on self-organizing feature map network cluster analysis, neural network is developed on the basis of a new non-supervised clustering method, analysis based on self-organizing feature map network clustering of the learning process, analyzed the weights of self-organization the process of neighborhood function and learning step-size problem of the general values are given based on the self-organizing feature map network clustering algorithm to achieve concrete and practical examples of tests, confirmed the correctness of the algorithm.

人工智能的教材啊- Artificial intelligence teaching material