PD_Fuzzy_controller

用MATLAB回归非线性模型参数,用MATLAB回归非线性模型参数(Linear regression model parameters with MATLAB)

本程序实现基于Haralick灰度共生矩阵的区域增长在matlab下实现(Implementation of this program based on Haralick GLCM achieved under the regional growth in matlab)

说明：  最优化设计有约束条件下的惩罚函数法的程序设计(Optimal design of constrained penalty function method of programming)

This file include matlab files to simulate example in "Basics of MATLAB and Beyond - Andrew Knight" e-book, you can use these .m file to improve your knowledge about matlab

一本讲解matlab的书籍，英文版，可以当做工作手册来用(learn matlab)

Matlab绘制子图函数的使用，Matlab绘制子图函数的使用(Matlab rendering the use of subgraph function, Matlab mapping function using subgraph)

说明：  MATLAB中plot函数的典型应用，适合初学者熟悉绘图函数(MATLAB plot function in a typical application, suitable for beginners familiar with the drawing functions)

说明：  主成分分析,matlab7.1上运行，很好的工具箱(PCA，component analysis matlab7.1 toolbox)

PLS - DN是一个有限的牛顿为解决非退化的分段线性系统的算法。 PLS - DN的展品可证明半迭代 财产即在全球范围内的精确解在有限数量的迭代算法收敛。被证明是收敛速度 至少前终止线性。 广泛的算法是在我们的AISTATS 2011纸描述：“一个非退化的分段线性系统的有限牛顿算法”。此演示包重新运行 在解决非负最小二乘法（NNLS）问题上的三个稀疏的设计矩阵，从哈威尔波音收集（达夫等人，1989年）第4.2节的实验。 随着PLS - DN，此演示包建议实施五 其他NNLS求解：LCP_Fisher_Newton，LCP_Lemke，PQN，TRESNEI，和SCD (PLS- DN is a finite non-degenerate Newton to solve the piecewise linear system algorithms. PLS- DN exhibits semi-iterative property to prove that the exact solutions worldwide in a limited number of iterative convergence. Proved to be terminated before the linear convergence rate at least. A wide range of algorithms in our AISTATS 2011 paper describes: " a non-degenerate sub-linear systems with limited Newton algorithm." This demo package to re-run in the settlement of non-negative least squares (NNLS) on the design of the three sparse matrices from the Harwell Boeing collection (Duff et al, 1989) Section 4.2 of the experiment. With PLS- DN, this demonstrates the implementation of five other NNLS package proposed solution: LCP_Fisher_Newton, LCP_Lemke, PQN, TRESNEI, and SCD)

[Numerical analysis]_ This algorithm computes the root of a function by the Wijngaarden_Dekker_Brent method.