This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Lecture 20 Duality Theory I UCSD Math 171A: Numerical Optimization Philip E. Gill http://ccom.ucsd.edu/~peg/math171 Friday, February 27th, 2009 Recap: Simplex method for standard form Solve B T = c B ; z N = c N N T ; ( z N ) s = min( z N ); if ( z N ) s then stop ; Solve Bp B = a s ; i = ( x B ) i ( p B ) i if ( p B ) i < 0; + if ( p B ) i 0; t = min { i } ; = t ; if = + then stop ; x B x B + p B ; ( x B ) t ; Exchange index t of B with index s of N ; UCSD Center for Computational Mathematics Slide 2/29, Friday, February 27th, 2009 Getting feasible How do we get a feasible basic solution of Ax = b ? Example: A = 2 3 1 1 1 2 1 1 ! , b = 5 3 ! UCSD Center for Computational Mathematics Slide 3/29, Friday, February 27th, 2009 2 x 1 + 3 x 2 + x 3 + x 4 = 5 x 1 + 2 x 2 + x 3 x 5 = 3 x 1 , x 2 , x 3 , x 4 , x 5 Add positive shifts x 6 and x 7 : 2 x 1 + 3 x 2 + x 3 + x 4 + x 6 = 5 x 1 + 2 x 2 + x 3 x 5 + x 7 = 3 x 1 , x 2 , x 3 , x 4 , x 5 , x 6 , x 7 The basis B = { 6 , 7 } defines the basic solution x B = 5 3 . UCSD Center for Computational Mathematics Slide 4/29, Friday, February 27th, 2009 Minimize the sum of infeasibilities minimize x R 7 x 6 + x 7 2 x 1 + 3 x 2 + x 3 + x 4 + x 6 = 5 x 1 + 2 x 2 + x 3 x 5 + x 7 = 3 x 1 , x 2 , x 3 , x 4 , x 5 , x 6 , x 7 If a feasible point exists, both x 6 and x 7 will be zero (i.e., nonbasic) at the end of phase 1. two other variables will be basic phase 1 solution is an initial feasible basic solution for phase 2 UCSD Center for Computational Mathematics Slide 5/29, Friday, February 27th, 2009 If the first constraint had been of the form 2 x 1 + 3 x 2 + x 3 + x 4 = 5 we would have included the shift x 6 as 2 x 1 + 3 x 2 + x 3 + x 4 x 6 = 5 but we still use + x 6 in the sum of infeasibilities. UCSD Center for Computational Mathematics Slide 6/29, Friday, February 27th, 2009 General case To find an initial vertex for Ax = b , x we define shifts v = x n +1 x n +2 . . . x n + m UCSD Center for Computational Mathematics Slide 7/29, Friday, February 27th, 2009 General case Define shifted constraints A V x v !...
View
Full
Document
This note was uploaded on 10/23/2010 for the course MATH 171a taught by Professor Staff during the Winter '08 term at UCSD.
 Winter '08
 staff
 Math

Click to edit the document details