搜索资源列表
biaoqing
- 对jaffe人脸库进行识别测试的主程序,将jaffe人脸库分为训练集和测试集两部分,首先对图片进行LBP+LPQ特征提取,然后svm分类识别,统计识别率-Jaffe face for the identification of the main test will jaffe face is divided into a training set and a test set of two parts, the first of LBP+LPQ image feature extractio
SVMwineclassification
- 选择178个样板的50%作为训练集,另外的作为测试集,用训练集对SVM进行训练得到分类模型,再用得到的分类模型对测试集进行类别标签预测。-Select 178 model 50 as the training set, the other as the test set, the SVM training classification model with the training set, and then the classification model by category label
FeatureSelection
- 为一个matlab的一个线性回归的例子,有测试集和训练集-For a matlab of a linear regression example, there are test sets and training sets
test2
- 为一个matlab的一个线性回归的例子,有测试集和训练集-For a matlab of a linear regression example, there are test sets and training sets
codes
- 手写数码(0到9)识别的神经网络框架。附有数据训练集和测试集。采用随机梯度下降的BP算法。可以修改参数,加入drop out和动量法。-Digital handwriting (0 to 9) of the neural network recognition framework. With training and test data sets. BP algorithm using stochastic gradient descent. Parameters can be modified
trbinpng-SVM
- 一个SVM训练工具,能够快速对训练集训练,给出分类判别函数-A SVM training tool, can quickly to training set training and classification discriminant function are given,
trginingtrainingfunction
- 一个SVM训练工具,能够快速对训练集训练,给出分类判别函数-A SVM training tool, can quickly to training set training and classification discriminant function are given,
mpca
- 以下Matlab项目包含用于模块化pca的源代码和Matlab示例。 代码有一些问题,orl面部从1.pgm命名为400.pgm.5从每个类中随机抽取的测试和训练集。图像分为4部分。-The following Matlab project contains the source code and Matlab examples used for modular pca. the code has some problem,orl faces are named 1.pgm to 400.pg
FilelistGenerator
- 用于CAFFE图片训练集、测试集的生成,可生成文件列表及分类,用于后续处理-Used for CAFFE training set and testing set generated images, can generate the file list and classification, for subsequent processing
chapter15_0
- svm 的参数优化,利用交叉验证法选择最优参数c g,最终提高训练集的分类准确率,更好的提高分类器性能-Svm parameter optimization, the use of cross-validation method to the optimal parameter c g, and ultimately improve the training set classification accuracy,better improve the classifier performan
chapter15_PSO
- svm 的参数优化,利用pso(粒子群优化算法)选择最优参数c g,最终提高训练集的分类准确率,更好的提高分类器性能-Svm parameter optimization, the use of pso (particle swarm optimization algorithm) to the optimal parameter c g, and ultimately improve the training set classification accuracy, better impr
chapter15_GA
- svm 的参数优化,利用ga(遗传优化算法)选择最优参数c g,最终提高训练集的分类准确率,更好的提高分类器性能-Svm parameter optimization, the use of ga (genetic optimization algorithm) to the optimal parameter c g, and ultimately improve the accuracy of the training set classification, better improve
gaSVMcgForClass
- svm 的参数优化,利用ga(遗传优化算法)选择最优参数c g,最终提高训练集的分类准确率,更好的提高分类器性能,这是ga的功能函数源码-Svm parameter optimization, the use of ga (genetic optimization algorithm) to the optimal parameter c g, and ultimately improve the training set classification accuracy, better imp
apd
- 一个SVM训练工具,能够快速对训练集训练,给出分类判别函数-A SVM training tool, can quickly to training set training and classification discriminant function are given,
模式识别第一次作业
- 1. 用 dataset1.txt 作为训练样本,用dataset2.txt 作为测试样本,采用身高和体重数据为特征,在正态分布假设下估计概率密度(只用训练样本),建立最小错误率贝叶斯分类器,写出所用的密度估计方法和得到的决策规则,将该分类器分别应用到训练集和测试集,考察训练错误率和测试错误率。将分类器应用到dataset3 上,考察测试错误率的情况。(1. using dataset1.txt as training samples as test samples by dataset2.tx
BP神经网络运动状态分类
- 该程序可以通过训练集对所构建的BP神经网络进行训练,并能通过测试集,即对不同的运动状态进行分类。(The program can train the constructed BP neural network through the training set, and can classify the different motion states through the test set.)
遗传算法的优化计算
- 遗传优化算法,自变量降维,训练集BP网络,单BP网络(genetic optimization algorithm)
Class_3_Code
- 将concrete_data.mat文件导入到MATLAB中,其中attributes为影响混凝土抗压强度的7个输入变量,strength为混凝土的抗压强度,即输出变量; 将整个数据集中的103个样本随机划分为训练集与测试集,其中训练集包含80个样本,测试集包含23个样本; 将训练集与测试集数据进行归一化; 建立BP神经网络,并训练; 利用训练好的BP神经网络对测试集中的23个样本的抗压强度进行预测; 输出结果并绘图(真实值与预测值对比图)(The concrete_data.mat
demo
- 利用训练集训练一个高斯模型,进行运动目标的提取(文件中包含数据集)(Use the training set to train a Gaussian model to extract the moving object (the file contains the data set))
FaceRec
- 人脸识别是一个有监督学习过程,首先利用训练集构造一个人脸模型,然后将测试集与训练集进行匹配,找到与之对应的训练集头像。最容易的方式是直接利用欧式距离计算测试集的每一幅图像与训练集的每一幅图像的距离,然后选择距离最近的图像作为识别的结果。这种直接计算距离的方式直观,但是有一个非常大的缺陷—计算量太大。如果每幅图像大小为100*100,训练集大小1000,则识别测试集中的一幅图像就需要1000*100*100的计算量,当测试集很大时,识别速度非常缓慢。(Face recognition is a s