Welcome![Sign In][Sign Up]
Location:
Search - perceptron weight

Search list

[matlabganzhi

Description:
Platform: | Size: 2048 | Author: wolf | Hits:

[AI-NN-PRnewp

Description: 利用感知器算法实现逻辑或和逻辑与功能,实现成功-Perceptron algorithm using the logical or and logical and functions, to achieve successful
Platform: | Size: 115712 | Author: 王新 | Hits:

[AI-NN-PRPerceptron-and-ADALINE-network

Description: 这些程序包括以下方面1.使用感知器和ADALINE网络对字母进行识别。2.随机选取初始权向量,选取适当的迭代步长(对ADALINE网络),用给出的四个输入训练样本,对上述两个网络分别进行训练,直到网络收敛;3.对Adaline网络选取不同的值,分别画出误差曲线,观察它们的变化规律;4.对感知器选取不同的初始权向量,分别计算每类训练样本到超平面的距离,观察它们的异同;5.训练结束后,检验网络的识别能力(使用100个检测样本,对应于每个取25个含噪的变形):6.比较Adaline和单神经元感知器的分类效果。-These programs include the following aspects of a perceptron and ADALINE networks to identify the letters. Randomly selected initial weight vector, select the appropriate step length (ADALINE network), four inputs with the given training sample, training on the above two networks, respectively, until the network convergence 3. Selected Adaline Network different values​ ​ , respectively, to draw the error curve, observe their variation . perceptron select the initial weight vector, calculated separately for each type of training samples to the hyperplane distance, to observe their similarities and differences 5. end of the training recognition ability of the test network (using the 100 samples tested, corresponds to each take 25 noisy deformation): 6. compare Adaline and single-neuron perceptron the classification results.
Platform: | Size: 19456 | Author: yuge | Hits:

[Graph RecognizePerceptron-Algorithm

Description: 模式识别是对样本进行聚类,感知器算法是通过迭代计算修正权向量,使样本满足条件,从而实现分类。本程序对感知器算法进行了改进,当权向量不满足时,立即退出此轮计算,进入下一轮迭代,从而减少了计算次数。程序对代码有详细注释,对样本的个数和维数自动判别。-Pattern recognition, clustering samples, perception algorithm is iterative correction weight vector samples meet the conditions, in order to achieve the classification. The program improved perception algorithm in power vectors are not met immediately exit the current round of calculation, proceed to the next round of iteration, thereby reducing the number of calculations. The program code has detailed notes, and automatically determine the number of samples and number of dimensions.
Platform: | Size: 2048 | Author: Mali Ang | Hits:

[Other1

Description: 双输入单输出系统 x1(1)=1 x2(1)=1 d(1)=1 x1(2)=-0.5 x2(1)=-1 d(1)=-1 x1(3)=3 x2(1)=1 d(1)=1 x1(4)=-2 x2(1)=-1 d(1)=-1 建立一个感知器网络,实现上述样本的分类,计算出相应的网络权值矩阵W-Dual-input single-output system x1 (1) = 1 x2 (1) = 1 d (1) = 1 x1 (2) =-0.5 x2 (1) =-1 d (1) =-1 x1 (3) = 3 x2 (1) = 1 d (1) = X1 (4) =-2 X2 (1) =-1 D (1) =-1 establish a perceptron network, and to achieve the classification of the above samples, calculate the corresponding network weight matrix W
Platform: | Size: 1024 | Author: chenkuan | Hits:

[matlabganzhiqi

Description: 感知器算法的基本思想是,对初始的或迭代中的增广权矢量 ,用已知的训练模式检验它的合理性,当不合理时,对其进行校正,校正方法实际上是最优化技术中的梯度下降法,上传的是用matlab解决感知器问题。-The Perceptron Algorithm The basic idea augmented weight vector of the initial iteration, the known training mode to test it reasonable, unreasonable, its correction, the correction method is actually the most optimization techniques the gradient descent method, upload matlab to solve perception problems.
Platform: | Size: 1024 | Author: 刘金 | Hits:

[AI-NN-PRganzhiqi2

Description: 3、单计算节点感知器,3个输入。给定3对训练样本对如下所示: X1 = (1,-2,0,-1) ,d1 =-1;X2 = (0,1.5,-0.5,-1),d2 = - 1;X3 = (-1,1,0.5,-1),d3 =1; 设初始权向量W(0)=(0.5,1,-1,0.5),2η=0.1。注意,x第四列为阈值恒等于-1,权向量中第后个分量为阈值,试根据以上学习规则训练该感知器。 -3, a single compute node sensor, three inputs. Given three pairs of training samples are as follows: X1 = (1,-2,0,-1), d1 =-1 X2 = (0,1.5,-0.5,-1), d2 =- 1 X3 = (-1,1,0.5,-1 ), d3 = 1 Let the initial weight vector W (0) = (0.5,1,-1,0.5), 2η = 0.1. Note, x fourth as the threshold constant equal to-1, right after the first two components of the vector is the threshold, try to train the perceptron learning rule based on the above.
Platform: | Size: 1024 | Author: 老三 | Hits:

[OtherHK

Description: HK算法思想很朴实,就是在最小均方误差准则下求得权矢量. 他相对于感知器算法的优点在于,他适用于线性可分和非线性可分得情况,对于线性可分的情况,给出最优权矢量,对于非线性可分得情况,能够判别出来,以退出迭代过程.(The idea of HK algorithm is very simple, which is to obtain the weight vector under the minimum mean square error criterion. Compared with the perceptron algorithm, it is suitable for linear separable and non-linear separable cases. For linear separable cases, the optimal weight vector is given. For non-linear separable cases, it can be distinguished to exit the iteration process.)
Platform: | Size: 17813504 | Author: mao12345 | Hits:

CodeBus www.codebus.net