vbemgmm

  在混合高斯模型参数估计方法上有很多方法，例如最大似然函数的EM算法，但是该算法容易出现过拟合，故本文提出了一个变分EM的算法来对参数进行估计，可以避免EM算法中的不足。 下面的示例文件中说明了使用下面的示例文件说明了用法 examplevbem，VBEM M示例文件 faithful.txt数据集为例(The parameters of Gauss mixture model estimation method has a lot of methods, such as the maximum likelihood function of EM algorithm, but the algorithm is prone to overfitting, so this paper proposes a variational EM algorithm to estimate the parameters, can avoid the shortcomings of the EM algorithm in the. The following example file illustrates the use of the following sample file Examplevbem, VBEM, M sample files Faithful.txt data set as an example)

vbemgmm\1.txt
vbemgmm\2.txt
vbemgmm\Alphawave.txt
vbemgmm\consist.asv
vbemgmm\consist.m
vbemgmm\dist2.asv
vbemgmm\dist2.m
vbemgmm\f1.txt
vbemgmm\faithful.txt
vbemgmm\faithful1.txt
vbemgmm\faithful2.txt
vbemgmm\faithful3.txt
vbemgmm\faithful4.txt
vbemgmm\faithful5.txt
vbemgmm\faithful分解.txt
vbemgmm\faithful分解2.txt
vbemgmm\gmm.asv
vbemgmm\gmm.m
vbemgmm\gmmactiv.asv
vbemgmm\gmmactiv.m
vbemgmm\gmmem.asv
vbemgmm\gmmem.m
vbemgmm\gmminit.asv
vbemgmm\gmminit.m
vbemgmm\gmmpak.asv
vbemgmm\gmmpak.m
vbemgmm\gmmpost.asv
vbemgmm\gmmpost.m
vbemgmm\gmmprob.asv
vbemgmm\gmmprob.m
vbemgmm\gmmsamp.m
vbemgmm\gmmunpak.m
vbemgmm\gmmVBEM.asv
vbemgmm\gmmVBEM.m
vbemgmm\gmmVBEMdemo.asv
vbemgmm\gmmVBEMdemo.m
vbemgmm\gmmVBEMorig.asv
vbemgmm\gmmVBEMorig.m
vbemgmm\kmeans.asv
vbemgmm\kmeans.m
vbemgmm\logsumexp.asv
vbemgmm\logsumexp.m
vbemgmm\mixGaussRnd.m
vbemgmm\MyEllipse.asv
vbemgmm\MyEllipse.m
vbemgmm\plotClass.m
vbemgmm\README.txt
vbemgmm\结果图
vbemgmm\结果图\Figure1.fig
vbemgmm\结果图\Figure4.fig
vbemgmm\结果图\gmmem1.fig
vbemgmm\结果图\untitled.fig
vbemgmm\结果图\untitled1.fig
vbemgmm\结果图\untitled3.fig
vbemgmm\结果图\untitled31.fig
vbemgmm\结果图\要
vbemgmm\结果图\要\500次的分类.fig
vbemgmm\结果图\要\500次的分类.jpg
vbemgmm\结果图\要\Figure2.fig
vbemgmm\结果图\要\Figure3.fig
vbemgmm\结果图\要\gmmem2.fig
vbemgmm\结果图\要\迭代次数图.fig
vbemgmm\运行界面结果.txt