GENERALIZED-TENSOR-

有关投影寻踪回归的matlab编程，其中运用了遗传算法对其参数进行优化，实现起来比较简单(The projection pursuit regression matlab programming, including the use of a genetic algorithm to optimize its parameters, simple realization)

pso程序的源代码 pso程序的源代码-粒子群优化算法源程序 作者：FashionXu 一个pso程序的源代码。在vc.net2003下面通过。 建议：看代码之前，请先弄明白pso是怎么回事。然后请对应着来：程序中用Agent代表一只鸟，PSO代表鸟群。阅读源代码，不要顺着看，先看main(),然后按照出现的东西的顺序，一个一个得来，呵呵，纯粹是建议。(program s source code pso pso program s source code - source code on Particle Swarm Optimization: FashionXu a pso program source code. In vc.net2003 following through. Recommendation: look at the code before you figure out how the matter pso. Then please correspond to: Program using Agent on behalf of a bird, PSO representative of the flock. Read the source code, do not look down, look at the main (), then follow the order of things there, one by one, come, Oh, merely suggestions.)

偏微分方程的求解，泊松方程的求解，一边大家学习交流(Partial differential equation, the solution of the Poisson equation, while we study the exchange)

教你如何做热图（heatmap），可以显示图像的百分比。(A mathod to tell you how to make a heatmap.)

用c编写的矩阵分解程序，对于不用 matlab的人来说，还是有一定用处(Prepared by c matrix factorization procedures, for people who do not have matlab, or there is some use)

说明：  协同过滤算法matlab代码,算法本身挺简单的，建议你自己试着编一下(Collaborative filtering algorithm matlab code )

给定一个样本集exset，样本数为M，每个样本点是N维向量，对于给定目标点d，d也为N维向量，要从exset中找出与d距离最近的k个点(k<=N)(knn code c++)

大致就是一个老鼠在一个六个边的迷宫（httpen.wikipedia.orgwikiRadial_arm_maze）里寻找食物（reward），然后找到后记忆下来 下一次去寻找的时候 可以更快的找到食物，具体要求和解释见附件，不过是英文的，里面还有2个example code也是关于(As a rat in a six side of the maze ( httpen.wikipedia.orgwikiRadial_arm_maze ) in search of food ( reward ), and then find the memory down next to find time can be faster to find food, specific requirements.zip)

brisk描述子,matlab语言，好懂易学(brisk descriptors, matlab language, easy to understand to learn)

　　编写一通用的预测法分析程序，要求有一定的错误处理能力，出错后能够使程序继续运行下去，直到分析过程结束。可通过不同的文法（通过数据表现）进行测试。(Write a generic prediction analysis procedure that requires a certain degree of error-handling capabilities, enabling the program to continue after an error to run it until the end of the analysis process. Can be tested by different grammar (by performance data).)