STAT 201B Lecture Notes - Lecture 1: Parametric Statistics, Parametric Model, Point Estimation

• School: Berkeley
• Department: Statistics
• Course: STAT 201B
• Professor: Elizabeth Purdom

Document Summary

- xn ) from a distribution f- cdf. Inference is the process of using data x to draw conclusion about f. A statistic is any function of data. Parametric model : set f can be described by numbers of parameters. F ~ ( n , 07 i =l . Poison , chi - square , normal , binomial ~ all parametric model f- for vector of parameters , using o as a subscript that cdf f depends on a. Nonparametric model infinite numbers of parameter fo cx ea) emphasize. Xnf ( smooth distribution) unknown , fixed constant. Interprets probability of long - term run. Focus on point estimation i ci , hypothesis. Posterior dist , update after seeing data tests. If n is sufficiently large, the probability en approaching o. Asymptotic normality z : n -7 d. Z n n co , t ) n , (cid:15482) no, i ) replace se ce n ) hold normality.