class06-3-handouts

class06-3-handouts - PSTAT 120B - Probability &...

Info iconThis preview shows pages 1–7. Sign up to view the full content.

View Full Document Right Arrow Icon
PSTAT 120B - Probability & Statistics Class # 06-3- Sufficiency Jarad Niemi University of California, Santa Barbara 7 May 2010 Jarad Niemi (UCSB) Sufficiency 7 May 2010 1 / 15
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Class overview Announcements Announcements Homework Homework 5 is up, due Monday 10 May by 4pm in SH 5521 Mid-term II Friday 14 May Covers 8.5–8.8,9.1–9.5 Jarad Niemi (UCSB) Sufficiency 7 May 2010 2 / 15
Background image of page 2
Class overview Goals (Minimal) sufficient statistics Rao-Blackwell Theorem Minimum-Variance Unbiased Estimators (MVUE) Jarad Niemi (UCSB) Sufficiency 7 May 2010 3 / 15
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Sufficient statistics Intuition Data Sufficient statistics Jarad Niemi (UCSB) Sufficiency 7 May 2010 4 / 15
Background image of page 4
Sufficient statistics Intuition Suppose X 1 , . . . , X n iid Ber ( p ). If we are interested in inference on p , what statistic do we need? What do I mean by inference ? Estimation, e.g. what is our best guess for p ? Confidence intervals, e.g. what is a reasonable range for p ? Hypothesis testing, e.g. is p = 0 . 5? What statistic do we need? Generally we need Y = n i =1 X i If we know Y , then individual X i are irrelevant Somehow Y sufficiently describes the data Jarad Niemi (UCSB) Sufficiency 7 May 2010 5 / 15
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Sufficient statistics Definition Definition Let Y 1 , Y 2 , . . . , Y n be iid from a probability distribution with unknown parameter θ . Then the statistic U = g ( Y 1 , Y 2 , . . . , Y n ) is said to be sufficient for θ if the conditional distribution of Y 1 , Y 2 , . . . , Y n given U does not depend on θ . Theorem If f Y ( y 1 , . . . , y n | θ ) is the joint pdf or pmf of the data and f U ( u | θ ) is the pdf or pmf of U = g ( Y 1 , Y 2 , . . . , Y n ) , then U is a sufficient statistic for θ if the ratio f Y ( y 1 , . . . , y n | θ ) /
Background image of page 6
Image of page 7
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 15

class06-3-handouts - PSTAT 120B - Probability &...

This preview shows document pages 1 - 7. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online