资源列表
cor_ls
- 把辨识分成两步进行:第一步:利用相关分析法获得对象的非参数模型(脉冲响应或相关函数);第二步:利用最小二乘法、辅助变量法或增广最小二乘法等,进一步求的对象的参数模型。如果模型噪声与输入无关,则Cor-ls相关最小二乘法(二步法)可以得到较好的辨识结果。Cor-ls相关最小二乘法(二步法)实质上是先对数据进行一次相关分析,滤除了有色噪声的影响,再利用最小二乘法必然就会改善辨识结果。能适应较宽广的噪声范围,计算量不大,初始值对辨识结果影响较小。但要求输入信号与噪声不相关-The identifica
yeRenAndChuanjiaoshi
- 人工智能中经典的野人与传教士问题,用Java实现-Savage and missionaries classic artificial intelligence problems, implemented in Java
RIV
- 适当选择辅助变量,使之满足相应条件,参数估计值就可以是无偏一致。估计辅助变量法的计算量与最小二乘法相当,但辨识效果却比最小二乘法好的多。尤其当噪声是有色的,而噪声的模型结构又不好确定时,增广最小二乘法和广义最小二乘法一般都不好直接应用,因为他们需要选用特定的模型结构,而辅助变量法不需要确定噪声的模型结构,因此辅助变量法就显得更为灵活,但辅助变量法不能同时获得噪声模型的参数估计。-Choose appropriate secondary variables, meet the relevant c
chapter3
- 清华大学人工智能课件第三章,仅供参考学习。内容翔实,方便自学。-Tsinghua University Artificial Intelligence courseware for reference only learning. Informative, convenient self-study.
DecisionTree
- 决策树算法。 数据写在Test文件里。 Attribute可以更改,以适应不同的要求。-The decision tree algorithm. The data is written in the Test file. Attribute can be changed to suit different requirements.
chapter2
- 清华大学人工智能课件,仅供参考学习。内容翔实,方便自学。-Tsinghua University Artificial Intelligence courseware for reference only learning. Informative, convenient self-study.
BP
- BP神经网络。 算法的接口很简单,只需修改Test文件,就可实现不同的功能。-BP neural network. Algorithm interface is very simple, just modify the Test file, you can achieve different functions.
nn
- 神经网络处理回归问题采用非批处理算法能够处理各种回归问题用c语言和矩阵运算解决回归问题-Neural network processing regression problems using non-batch algorithm to be able to handle all kinds of regression problems to solve regression problems with c language and matrix operations
flexibles
- 这是一个加工车间的优化调度的算法,算法应用矩阵编码的方式,实现优化求解.制造是一种十分重要的加工生产模式,本程序模拟了一条生产线.-This is a flexible processing workshops optimal scheduling of the genetic algorithm, the algorithm application matrix coding modes to achieve optimization, and welcome other authors as
SimulatedAnnealing
- 模拟退火算法(Simulated Annealing,简称SA算法)是模拟加热熔化的金属的退火过程,来寻找全局最优解的有效方法之一。-Simulated annealing algorithm (Simulated Annealing, SA algorithm) annealing process is simulated heating molten metal, one of the effective way to find the global optimum solution.
The-horse-around-algorithm
- 用回朔法实现N*N棋盘的周游,每个格子经过且仅经过一次-N* N chessboard backtracking traveling, each grid after and only after a
RFM
- 辨识所使用的数据长度保持不变,每增加一个新数据就抛掉一个老数据,使参数估计值始终只依赖于有限个新数据所提供的新消息,克服了遗忘因子法不管多老的数据都在起作用的缺点,因此该算法更能有效的克服数据饱和现象。-Identify the use of data length remain the same, every time you add a new data will throw away an old data, make the parameter estimate always depen
