人工智能模式识别中基于非负矩阵分解生成特征空间的算法

本系统是一个有关本体的专家知识库系统，目的是实现专家知识的语义查询、推理。-This system is an expert on ontological knowledge base system, aimed at achieving the expert knowledge of the semantic query and reasoning.

使用labview编写的神经网络程序，多输入多数出。-prepared by the use of PC neural network programming, multi-input from the majority.

Kohonen网络的学习过程可描述为：对于每一个网络的输入，只调整一部分权值，使权向量更接近或更偏离输入矢量，这一调整过程就是竞争学习。随着不断的学习过程，所有输入矢量都在输入矢量空间相互分离，形成了各自代表输入空间的一类模式，这就是Kohonen网络的特征自动识别的聚类功能。请解压缩后按照readme提示进行操作。-Kohonen network learning process can be described as follows : for each one network input, only part of adjusting the weights, weight vector closer to or further from the input vector, the adjustment process is competitive learning. With the continuous process of learning, all the input vector in the importation of vector space separate form their own representatives importation of a type of space, and this is a Kohonen network of automatic clustering functions. Please decompress after readme according to instructions.

SVM的讲义！具体的内容可以自己看！很详细的-SVM lectures! Specific content can look! Very detailed

小波人工神经网络C语言实现(代码)，以及相关支持库，已调试，可运行．-Wavelet artificial neural network C language (code), and related support libraries, have been debug run.

仿真1：首先把网络温度参数T固定在100，按工作规则共进行1000次状态更新,把这1000次状态转移中网络中的各个状态出现的次数Si（i＝1，2，…，16）记录下来 按下式计算各个状态出现的实际频率： Pi=Si/∑i=1,NSi=Si/M 同时按照Bo1tzmann分布计算网络各个状态出现概率的理论值： Q（Ei）=（1/Z）exp（-Ei/T） 仿真2：实施降温方案，重新计算 采用快速降温方案：T（t）= T0/（1+t） T从1000降到0.01，按工作规则更新网络状态 当T=0.01时结束降温，再让T保持在0.01进行1000次状态转移，比较两种概率-a simulation : First of all network parameters temperature T fixed at 100 and, according to the rules for a total of 1000 to update the state, this state of the 1000 network transfer of all states for the number of Si (i = 1, 2, ..., 16) all recorded determined by the formula state-of the actual frequency : Pi = Si/i = 1, NSi = Si/M in accordance with Bo1tzmann distributed computing network of states all probability the theoretical value : Q (Ei) = (1/Z) exp (- Ei/T) Simulation 2 : implementation of cooling, re-using rapid cooling programs : T (t) = T0/(1 t) T dropped to 0.01 from 1000 and, according to the rules updated network state when T = 0.01 at the end of cooling, let T at 0.01 for the 1000 state tran

3个径向基网络的matlab源程序[matlab],一维输入，一维输出，逼近效果很好！-three RBF network Matlab source [Matlab], one-dimensional input and one-dimensional output. Approximation good effect!

人工智能里的A-star算法,用于机器人的路径规划和寻优.-AI"s A-star algorithm for robot path planning and optimization.

遗传算法解决双变量的函数最优化问题，有按钮的界面，用bc所编，生动模拟遗传进化过程-genetic algorithm to solve the two- variable optimization function, the button interface, using bc prepared by the vivid simulation of the process of genetic evolution

C# 15数码算法，A star 函数实现 -C# 15 Digital Algorithm