学习排序问题的一种方法是将问题转化为分类问题。

一个更精度的平滑涵数, 可用于股票交易系统.用于Amibroker 平台-a more accurate smooth average function for amibroker trading system

EM algorithm with a Rauch-Tung-Striebel smoother and an M step,内有说明-EM algorithm with a Rauch-Tung-Striebel"s moother and an M step, which is described

我们解决八数码问题,广度优先搜索可能会导致内存不够用,采用启发式搜索,启发函数为f(x)=g(x)+h(x) g(x)为该结点不同于目标结点的个数,h(x)为该结点的深度,选择那f(x)结点最小的那个结点进行扩展, 引入了一个"扩展数组[4]"(因为扩展的结点最多只有4个),该数组保存着某个结点的0点向各个方向的扩展的结点对象,然后对该扩展数组进行分析,利用启发函数在进行递归扩展... -us solve eight digital, BFS may lead to insufficient memory, using heuristic search, heuristic function f (x) = g (x) h (x) g (x) for the target node is different from the number of nodes, h (x) for the nodes of depth, the choice between f (x) nodes that the smallest node expansion. the introduction of an "expanded array [4]" (because the expansion of the nodes at most four), The array maintains a node of 0:00 a.m. to expand in all directions in the target node, and then expanded array analysis, use heuristic function during recursive expansion ...

自适应遗传算法 最基础的版本需要可以自己修改目标函数-Adaptive genetic algorithm version of the most basic needs can modify the objective function

基于机器学习的自动逻辑推理机。本程序用归结反演策略实现了命题逻辑系统的自动推理。把要解决的问题作为一个要证明的命题，其目标公式被否定并化成子句形，然后添加到命题公式集中去,把消解反演系统应用于联合集，并推导出一个空子句(NIL)，产生一个矛盾。-based on the automatic machine learning logical reasoning machine. The procedures used attributed inversion strategy to achieve the propositional logic of automatic reasoning. Put the problem to be solved as a proposition to prove, with the goal being negated formula and shape into a phrase and then added to the Proposition formula to concentrate, digestion system for joint inversion sets and derived an empty phrase (NIL), resulting in a contradiction.

模式识别一份很好的作业,包括线性分类器；最小风险贝叶斯分类器；监督学习法分层聚类分析；K－L变换提取有效特征,支持向量机-a very good operation, including linear classification; Minimum risk Bayesian classifier; Supervised learning method Hierarchical clustering analysis; K-L transform effective features, support vector machines

使用labview编写的dice游戏，很有趣味性。-prepared by the use of PC dice games, very interesting.

微分进化算法DE是1995年由Rainer Storn和Kenneth Price首先提出。DE已被证明在求解过程中具有高效性、收敛性、鲁棒性等优点[5,6] 。它在许多优化问题中都表现出优于自适应模拟退火算法、POS 算法、GA算法的性能。DE利用实数值参数向量作为每一代的种群，它的自参考种群繁殖方案与其他优化算法不同。-DE was first proposed in 1995 by Rainer Storn and Kenneth Price and had been proven that it is efficient, convergent, and robust in the solution process. [5,6] In many optimize questions, its performance surpasses the adaptive simulated annealing algorithm, the POS algorithm, and the GA algorithm. DE gets each generation of population using the real value parameter vector, and the crucial idea of DE is a scheme for generating trial parameter vectors.

基于Web页面的人工智能控件例证 [VB源码]-Web page based on the example of artificial intelligence control [VB code]