附加了动量项BP神经网络

一个遗传算法源程序 可以作为学习改进使用 希望对大家有所帮助-A genetic algorithm source code can be used as learning to improve the use of want to be helpful to everyone

eview 编写的时间序列预报程序 可以用于金融预报，等-eview the preparation of time-series forecasting procedures can be used for financial forecasting, etc.

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.

关于神经网络入门学习的教程，适用于刚开始学习的人-Getting Started on the neural network learning tutorial for people who just started learning

学习魔方编程技巧，仅供参考，有信息请反馈。-learning programming skills, for reference purposes only, with information feedback. Thank you

本程序是用c语言编写的基于决策树分类方法的数据挖掘算法，它对测试集进行分类，挖掘出潜在的规则-this procedure is used to prepare the language c decision tree classification based on the data mining algorithm, which tests set for classification, tapping the potential rules

很好的边缘检测源代码，使用vc++6.0编写，主要用于医学图像的边缘检测-good source edge detection, the use of vc 6.0 preparation, mainly for Medical Image Edge Detection

一个可逆跳转蒙特卡罗采样（RJMCMC）算法详细程序，内附相关论文，对照论文看算法，便于理解。包含多种运动方式（增加，减少，分裂，合成，更新）-A Reversible Jump Monte Carlo sampling (RJMCMC) algorithm detailed procedures, enclosing the relevant papers, watch the control thesis algorithm, easy to understand. Contains a wide range of movement (increase, decrease, fragmentation, synthesis, update)

利用 免疫算法求解TSP问题 可以精确求出50个城市-Using immune algorithm for solving TSP problems can be accurately calculated 50 cities

演化神经网络拓扑结构编写的扫雷机器人程序-evolution of neural network topology structure of the mine-sweeping robot procedures