Anskey3

Anskey3 - x i )] Second derivatives H = 2 4 @ 2 L (...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
Answer Key for Homework 3 1. Report descriptive statistics for this data set. What is the fraction of duration that are censored? duration d age educ white rate min 8 0 17 9 0 1 max 1174 1 33 17 1 6 mean 229.3 0.34 22.52 12 .5 3.58 median 140 0 22 12 .5 3 std 256.2 .48 3.52 1.43 051 1.25 34% is censored 2. Compute the MLE of & 0 if 1 = 0 . Do this with and without accounting for censored observations. Without accounting for censored observations Log likelihood function is L ( ) = n X i =1 ln f ( y j x ) = n X i =1 [( & 0 + 1 x i ) ± y i exp( & 0 + 1 x i )] = n X i =1 [ & 0 ± y i exp( & 0 )] From @L ( ) @& 0 = 0 ; ^ & 0 = ln 0 @ n X y i 1 A = ± 5 : 435 With accounting for censored observations L ( ) = n X i =1 [(1 ± d i ) ln f ( y i j x ) + S ( y j x )] ; where d = 1 if censored = n X i =1 [(1 ± d i )( & 0 + 1 x i ) ± y i exp( & 0 + 1 x i )] = n X i =1 [(1 ± d i ) & 0 ± y i exp( & 0 )] (*) From @L ( ) @& 0 = 0 ; ^ & 0 = ln 0 @ X (1 ± d i ) X y i 1 A = ± 6 : 514 1
Background image of page 1

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

View Full DocumentRight Arrow Icon
3. Show that the log likelihood function at & 0 = 1 = 0 is equal to -11465. Remember to take account of the censoring At & 0 = 1 = 0 ; From (*), L ( ) = X y i 4. L ( ) = n X i =1 [(1 d i )( & 0 + 1 x i ) y i exp( & 0 + 1 x i )] @L ( ) @& 0 = n X i =1 [(1 d i ) y i exp( & 0 + 1 x i )] @L ( ) 1 = n X i =1 [(1 d i ) x i x i y i exp( & 0 + 1
Background image of page 2
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: x i )] Second derivatives H = 2 4 @ 2 L ( & ) @& 2 @ 2 L ( & ) @& & 1 @ 2 L ( & ) @& 1 & @ 2 L ( & ) @& 2 1 3 5 = 2 6 6 6 6 4 & n X i =1 [ y i exp( & + & 1 x i )] & n X i =1 [ x i y i exp( & + & 1 x i )] & n X i =1 [ x i y i exp( & + & 1 x i )] & n X i =1 & x 2 i y i exp( & + & 1 x i ) 3 7 7 7 7 5 5. Maximize the log likelihood function using the Newton-Raphson algorithm with anlytic derivatives. Use zeros as starting values. Also use the MLE for & obtained above and & 1 = 0 : ^ & mle = ( & 5 : 6139 ; & : 0198) 6. Compute the standard errors of the MLE. AV ar ( & ) = ( & E [ H ]) & 1 =n standard errors = (1 : 349 ; : 112) 2...
View Full Document

This note was uploaded on 07/22/2008 for the course ECON 513 taught by Professor Rashidian during the Fall '07 term at USC.

Page1 / 2

Anskey3 - x i )] Second derivatives H = 2 4 @ 2 L (...

This preview shows document pages 1 - 2. Sign up to view the full document.

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