MA05425: 统计建模
R-project (CRAN)   本地下载R2.7.2

考核方式: 考试(50%) + 作业(30%) + project(20%)

教材：

• A.C. Davison, Statistical Models, Cambridge University Press, 2003
  SMPracticals:本书中例子和练习。 安装R package "SMPracticals" (含数据集，和一些用到的函数)。 作者主页
  

  课本中的数据集列表

内容

讲稿	摘要	补充材料	作业
9.8:     Introduction	模型= 分布+系统变差	EM algorithm	作业1
9.22:   Variation and uncertainty	分布及其刻画
9.27:   Likelihood (1)	似然=Pr(data),information		作业2: P115, problem 6.
10.6:   Likelihood (2)	Profile likelihood	需加载MASS: library(MASS)
10.13: Likelihood (3)	Irregularity		作业3:P149: 1. P157-159: 4,8,10,11,12
10.20: Likelihood (4)	Model selection, EM	Information theory AIC
10.27: Likelihood (5)	Mixture, image processing	演示	作业4	答案
11.3:   Stochastic models	Markov chain		P254: 1	答案
11.10: Markov random fields	Hammersley-Clifford theorem	Image denosing	作业5: p255:12
11.17: Gaussian graphical model	&Omega= covariance; ℧=&Omega-1(precision matrix) - &Omegaij=0: indepdent of i,j; - ℧ij=0: conditionally indepdent of i,j;		作业6(全部选做): P294-296: 2,3,7
11.24: State-space model	Hidden Markov model (HMM)	HMM的EM算法	作业7: analyze fetus data ( help)
12.1:   Volatility models			Project
12.8:   Monte Carlo	MCMC
12.15: Gibbs sampler