gaot

AR模形拟合工具，提供AR模型的拟合与模拟(AR model )

levy法编制辨识程序，基本精确辨识了系统各个参数，因为误差准则的原因，系数略有差别。(Levy method prepare identification procedure, basic accurate identification of each parameter of the system, because the cause of the error criterion, a slightly different coefficient.)

报童问题的计算机仿真 ％tm一轮实验的预定模拟天数 ％t一轮实验的仿真天数累积值 ％z订报量 ％z 最优订报量 ％g订报量z之上界 ％s1损失值之累计值 ％s最小损失值值 ％r按概率分布产生随机售报量样本(newsboy problem of computer simulation% tm an experimental simulation of the target number of days a t% of the experimental days of accrued Simulation plot value volume%% z z Factory Workers optimal quantity% g subscription volume z% above the industry s1 loss value% cumulative value of the smallest losses's value % r value of the probability distribution of random reported sales volume samples)

这是个人总结的各种聚类算法的优缺点比较，包括了常用的一些聚类算法，如层次聚类算法、分割聚类算法、基于约束的算法等(The comparison of various clustering algorithms by myself,including 5 common clustering algorithms)

Total program for radial system

说明：  标准粒子群优化算法,适用于matlab。方便实用(Standard particle swarm optimization for matlab. Convenient and practical)

code for neural network classifier

K均值算法描述： 给定类的个数K，将N个对象分到K个类中去， 使得类内对象之间的相似性最大，而类之间的相似性最小。(K-means algorithm described: a given number of classes K and N objects assigned to the K-th class, so that objects within the class of the similarity between the maximum, while the similarity between the minimum class.)

用matlab的bayes方法实现系统辨识(Bayes methods using matlab realize the system identification)

微软的一道题目，有12个外表一模一样的小球，其中有一个的质量和其他11个球不一样（11个球的质量完全相同），而且该球不知道是轻点还是重点。现在给你一个天平（无砝码），只有3次测量机会，请你找出该球。球号从a到l(小写)(Microsoft' s one topic, there are 12 look exactly the same ball, one of which the quality and not the same as the other 11 balls (11 balls of the same quality), and that the ball does not know or focus of light spots. Now to give you a balance (no weight), only three measurements chance, would you please identify the ball. Ball number from a to l (lowercase))