资源简介
PCA实现图像分类,论文+代码+测试/训练样品,华工出品,必属精品。
代码片段和文件信息
function [distanceLdistanceGLogLikelihood]=apply_model(TLmeanGmeanCgppcainvCgppcaVtot)
% This function Apply_Model uses the matrices and means from train_model.m
% to calculate the mahalanobis distance to global and local feature
% training data set which are combined to a log Likelihood ratio.
%
% [distanceLdistanceGLogLikelihood]=apply_model(TLmeanGmeanCgppca invCgppcaVtot)
%
% inputs
% T: Matrix with all feature vectors from the test data first
% feature vector T(:1)
% G: Matrix with all feature vectors from the global (not) training data
% inputs (from train_model.m)
% Lmean : The mean of the local feature vectors
% Gmean : The mean of the global feature vectors
% Cgppca : Covariance matrix after 3 PCA steps
% invCgppca : Inverse Covariance matrix after 3 PCA steps
% Vtot : The Rotation matrix of the 3 PCA steps% outputs
%
% outputs
% distanceL : The distance to the local training data set
% distanceG : The distance to the global training data set
% LogLikelihood : The log likelihood of test feature vector to belong to
% the local set.
%
% In the 3 PCA steps the feature vectors are rotated in a way that
% the Mahalanobis distance to the local set can be deteremined by
% FeatureA‘*FeatureA and to the global set by FeatureA‘*invCgppca*FeatureA.
% FeatureA is the mean substracted and rotated FeatureVector of a certain
% test (image/coordinate).
%
% Literature : Kroon D.J. and van Oort E.S.B. and Slump C.H. “Multiple
% Sclerosis Detection in Multispectral Magnetic Resonance Images with
% Principal Components Analysis“
%
% Function is written by D.Kroon University of Twente (July 2009)
distanceL=zeros(1size(T2));
distanceG=zeros(1size(T2));
for i=1:size(T2)
u=Vtot‘*(T(:i)-Gmean);
v=Vtot‘*(T(:i)-Lmean);
distanceG(i)=(u‘*invCgppca*u);
distanceL(i)=(v‘*v);
end
offset=log(det(Cgppca));
LogLikelihood=0.5*(distanceG-distanceL)+0.5*offset;
属性 大小 日期 时间 名称
----------- --------- ---------- ----- ----
文件 2019 2009-07-21 20:18 PCA_classifier_version1b\apply_model.m
文件 4801 2010-02-11 20:01 PCA_classifier_version1b\example_classifier_ms.m
文件 3692 2009-07-21 20:02 PCA_classifier_version1b\get_feature_vectors.c
文件 542 2009-07-21 18:50 PCA_classifier_version1b\get_feature_vectors.m
文件 742746 2008-08-13 18:05 PCA_classifier_version1b\Literature\MICCAI_MS_Challenge_UTwente_Final.pdf
文件 42372 2009-07-21 20:47 PCA_classifier_version1b\TestData\patient3_FLAIR.png
文件 51434 2009-07-21 17:26 PCA_classifier_version1b\TestData\patient3_T1.png
文件 84151 2009-07-21 17:27 PCA_classifier_version1b\TestData\patient3_T2.png
文件 3865 2010-02-11 19:56 PCA_classifier_version1b\train_model.m
文件 69519 2009-07-21 17:09 PCA_classifier_version1b\TrainingData\patient1_FLAIR.png
文件 739 2009-07-21 17:22 PCA_classifier_version1b\TrainingData\patient1_lesion.png
文件 70981 2009-07-21 17:21 PCA_classifier_version1b\TrainingData\patient1_T1.png
文件 82651 2009-07-21 17:22 PCA_classifier_version1b\TrainingData\patient1_T2.png
文件 60989 2009-07-21 20:48 PCA_classifier_version1b\TrainingData\patient2_FLAIR.png
文件 577 2009-07-21 17:25 PCA_classifier_version1b\TrainingData\patient2_lesion.png
文件 68764 2009-07-21 17:24 PCA_classifier_version1b\TrainingData\patient2_T1.png
文件 82603 2009-07-21 17:25 PCA_classifier_version1b\TrainingData\patient2_T2.png
文件 1314 2014-02-12 13:04 license.txt
相关资源
- Spark机器学习 (彭特里思著) 中文
- EM&GMM.zip
- 数据挖掘、机器学习在客户细分中应
- 车牌识别2400个正样本
- machine-learning-ex4
- PRML-Solutions to Exercises Tutors Edition
- 斯坦福大学 2014 机器学习教程中文笔
- 重庆大学软件学院机器学习ppt
- 斯坦福大学 2014 机器学习教程个人笔
- Bayesian Networks With Examples in R280858
- 吴恩达机器学习课件
- 张志华《机器学习导论》笔记和补充
- 《机器学习〔中文版〕》 PDF
- High-Dimensional Probability: An Introduction
- 机器学习那些事
- 机器学习十大算法C实现
- Dence CRF 条件随机场图像分割
- 基于深度神经网络的用户会话推荐算
- 上海交通大学-张志华-机器学习/统计
- Introduction to Machine Learning Ethem ALPAYDI
- 黄广斌-极限学习机讲义
- 斯坦福机器学习ML公开课笔记1-15完整
- Spark机器学习高清文字版/可复制/带目
- 比赛测试数据3
- Machine Learning_Algorithms and Applications (
- 基于Spark大数据电商推荐系统源码
- 人工智能导论全套课件.ppt
- 数据挖掘与机器学习课设
- XGBOOST算法原理及应用介绍
- CS231N作业源码,斯坦福机器学习课
评论
共有 条评论