定向图形模型Directed Graphical Models |
|
课程网址: | http://videolectures.net/aibootcamp2011_archambeau_cgm/ |
主讲教师: | Cedric Archambeau |
开课单位: | 伦敦大学学院 |
开课时间: | 2011-03-31 |
课程语种: | 英语 |
中文简介: | 在本文中,我将介绍定向图形模型的基本概念。然后介绍了EM算法并讨论了隐变量模型中的学习问题,考虑了几种混合模型(离散隐变量)、概率PCA(连续隐变量)和扩展。接下来,我描述回归的条件模型,绘制到最小二乘和岭回归的链接。最后,介绍了高斯过程回归。 |
课程简介: | In this talk I introduce the basic concepts of directed graphical models. I then introduce the EM algorithm and discuss learning in latent variable models, considering several mixture models (discrete latent variables), probabilistic PCA (continuous latent variables) and extensions. Next, I describe conditional models for regression, draw links to least squares and ridge regression. Finally, the talk is ended with an introduction to Gaussian process regression. |
关 键 词: | 向图形模型; 离散潜变量; 连续潜变量; 高斯过程 |
课程来源: | 视频讲座网 |
最后编审: | 2019-12-19:lxf |
阅读次数: | 48 |