搜索资源列表
11
- 神经网络实例集。包括以下几个程序单层线性神经网络实例、感知器神经元解决较复杂输入向量的分类问题、基于感知器神经网络处理复杂的分类问题、数值分析程序matlab-GUI、用BP网络完成函数的逼近源程序、自组织特征映射应用实例-Examples of neural network sets. Procedures include the following examples of single-layer linear neural network, perceptron neuron input
TuXiangShiBie
- 该软件需用Delphi7设计，采用灰度共生矩阵的方法对肝脏超声图像进行纹理特征提取。通过神经网络进行分类处理。-The software required Delphi7 design, the use of gray-scale co-occurrence matrix method of liver ultrasound image texture feature extraction. Through the neural network classification.
Ncut_SVM
- 此源码可对图像进行Ncut分割，并且集成了特征提取和SVM分类的功能。-This source can be Ncut image segmentation, and integrates the feature extraction and SVM classification function.
zifushibie
- 有数字1和2的图片各四十张，利用mat lab读取图片并将其二值化，得到只由0和1表示的矩阵，根据数字的不同特征，选择两个特征值，用来对分类器进行训练，最后用剩下的5个1和5个2来进行测试，看分类器是否可以正确的对数据进行分类。-Figures 1 and 2 pictures of the 40, using mat lab picture and secondly, the value of reading, and received only by 0 and 1, said matrix
basedonbayes
- 空间数据分析中最常用的是聚类分析，基于bayes的统计分析识别，一种特征分类方法-Spatial data analysis is the most commonly used cluster analysis, based on the Bayes statistical analysis of recognition, a feature classification method
tezhengxuanzhe
- 利用最小互信息实现向量的特征选择，优化分类器的设计，原创-The use of mutual information to achieve the smallest feature selection vectors, optimizing the classifier design, originality
simpleABdemo
- Adaboost算法的基本思想是:利用大量的分类能力一般的弱分类器(weaker ifier)通过一定的方法叠加(boost)起来，构成一个分类能力很强的强分类器 眼eClassifier)，再将若干个强分类器串联成为分级分类器(ClassifierCaseade) 图像搜索检测。本文就是利用Adaboost算法将由类haar特征生成的弱分类器 成为强分类器，再将强分类器串联成为分级分类器。 -Adaboost algorithm basic idea is: the abi
KL2
- 本程序利用K-L变换已经K-L变换的最优压缩，建立分类器，并选择投影方向，画出投影过后的效果-This procedure has been the use of KL transform KL transform optimal compression, the establishment of classifier, and choose the direction of projection, drawn after the effect of projection
mtl4-alpha-1-r6418
- 矩阵运算源码最新版本，支持矩阵乘法、转置求逆，特征值特征向量等操作。-Source the latest version of matrix operations in support of matrix multiplication, transpose inverse, eigenvalue eigenvector and so on.
include
- 一种基于字符骨架中封闭曲线特征和纵向线条特征的两级初分类算法,将原来含有62个元素的待识别字符集比较平均的分成了三个子集,降低了后续处理的难度。-Research on Printed Character Recognition System Based on BP Neutral Network
Wavelet_Based_Feature_Extraction_for_SVM_for_Scree
- 支持向量机在模式识别和分类中应用广泛， 小波方法的多尺度特性也众所周知。 本文将小波和支持向量机相互结合实现特征提取。-Support vector machine in pattern recognition and classification of the application of a wide range of multi-scale wavelet method is also well-known characteristics. In this paper, wavel
Selforg
- 自组织特征映射网络进行图像分类识别（神经网络实用教程）-Self-organizing feature map network image classification Recognition [Neural Network Practical Guide]
lpccfeatureextraction
- 倒谱系数作为特征提取方法，数据采用的是麻省理工的PAF数据，提取出的特征属性具有较好的分类性能-Cepstral coefficients as feature extraction methods, data used by the PAF are the Massachusetts data, extracted features of the property has good classification performance
pattern-recognition-simulation
- 用mushrooms数据对模式识别课程讲述的各种模式分类方法[线性分类，Bayesian分类，Parzen窗，KNN]和特征选择和降维方法[PCA,LDA]进行了模拟，并给出了各类分类方法的结果，-It s the simulations about linear classification ,Bayesian ,Parzen and KNN of pattern recognition .And ,It gives the results.
PCA
- 对输入的高维特征向量进行pca降维后输出低维的特征向量-PCA dimensionality reduction
ImageTest6
- 数字图像处理学--VC++实现 边缘检测 特征提取 图像识别 分类算法-Visual C++ Digital Image Proccessing
m10_9
- 一维自组织特征映射网络对输入向量空间进行识别分类-One-dimensional self-organizing feature map network input vector space to identify categories
TuXiangShu
- 本书主要论述了智能图像处理技术,系统介绍了智能图像处理技术的有代表性的思想、算法与应用，跟踪了图像处理技术的发展前沿。全书共分为15章，重点讨论了图像边缘检测、图像分割、图像特征分析、图像配准、图像融合、图像分类、图像识别、基于内容的图像检索与图像数字水印。此外，为了内容的完整性，本书还介绍了图像预处理技术，如图像采集、图像变换、图像增强、图像恢复、图像编码与压缩-This book focuses on intelligent image processing technologies, sy
WeightedFeature
- 给出两个加权特征，一个是加权笔画密度特征，另外一个是加权外围特征，用一级汉字实验结果表明，这两个特征具有很强的汉字信息，能很好的为模式分类提供有效的特征- Give out two weighted feature abstraction method One is the weighted stroke density feature , the other is the weighted Periphery feature.The reslut of experiment on the f
TextMining
- 介绍自动文本分类的一个ppt，详细的介绍了自动文本分类的特征提取，分类算法以及评估。-Introduced an automatic text classification ppt, a detailed introduction of automatic text categorization of feature extraction, classification algorithms, as well as assessment.