# APME_6 - Applied Probability Methods for Engineers Slide...

This preview shows pages 1–11. Sign up to view the full content.

Click to edit Master subtitle style Applied Probability Methods for Engineers Slide Set 6

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

View Full Document
Click to edit Master subtitle style Chapter 13 of Winston’s Operations Research Book Decision Making under Uncertainty
Newsvendor Example n A newsvendor sells newspapers each day n Cost of a paper to vendor is 20¢ n Revenue from selling a paper is 25¢ n Unsold papers are worthless n Sales are uniform between 6 and 10 (S = {6, 7, 8, 9, 10}) n If i papers are purchased by vendor, and j papers are demanded, profit is rij, where n rij = 25i – 20i = 5i, if i ≤ j n rij = 25j – 20 i, if i ≥ j n What should the vendor do?

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

View Full Document
Table of rewards n Why did we not consider ordering 0 – 5 papers? Papers Demanded Papers ordered 6 7 8 9 10 6 30 ¢ 30 ¢ 30 ¢ 30 ¢ 30 ¢ 7 10 ¢ 35 ¢ 35 ¢ 35 ¢ 35 ¢ 8 - 10 ¢ 15 ¢ 40 ¢ 40 ¢ 40 ¢ 9 - 30 ¢ - 5 ¢ 20 ¢ 45 ¢ 45 ¢ 10 - 50 ¢ - 25 ¢ 0 ¢ 25 ¢ 50 ¢
Maximin Criterion n Chooses action ai with largest value of Papers Demanded Papers ordered 6 7 8 9 10 6 30 ¢ 30 ¢ 30 ¢ 30 ¢ 30 ¢ 7 10 ¢ 35 ¢ 35 ¢ 35 ¢ 35 ¢ 8 - 10 ¢ 15 ¢ 40 ¢ 40 ¢ 40 ¢ 9 - 30 ¢ - 5 ¢ 20 ¢ 45 ¢ 45 ¢ 10 - 50 ¢ - 25 ¢ 0 ¢ 25 ¢ 50 ¢ Maximin

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

View Full Document
Maximax Criterion n Chooses action ai with largest value of Papers Demanded Papers ordered 6 7 8 9 10 6 30 ¢ 30 ¢ 30 ¢ 30 ¢ 30 ¢ 7 10 ¢ 35 ¢ 35 ¢ 35 ¢ 35 ¢ 8 - 10 ¢ 15 ¢ 40 ¢ 40 ¢ 40 ¢ 9 - 30 ¢ - 5 ¢ 20 ¢ 45 ¢ 45 ¢ 10 - 50 ¢ - 25 ¢ 0 ¢ 25 ¢ 50 ¢ Maximax
Minimax Regret n For each possible state sj, find action i*(j) that maximizes rij n For any action ai and state sj, opportunity loss (or regret) is ri*(j),j – rij n Subtract from max column value Papers Demanded Papers ordered 6 7 8 9 10 6 30 ¢ (0) 30 ¢ (5) 30 ¢ (10) 30 ¢ (15) 30 ¢ (20) 7 10 ¢ (20) 35 ¢ (0) 35 ¢ (5) 35 ¢ (10) 35 ¢ (15) 8 - 10 ¢ (40) 15 ¢ (20) 40 ¢ (0) 40 ¢ (5) 40 ¢ (10) 9 - 30 ¢ (60) - 5 ¢ (40) 20 ¢ (20) 45 ¢ (0) 45 ¢ (5) 10 - 50 ¢ (80) - 25 ¢ (60) 0 ¢ (40) 25 ¢ (20) 50 ¢ (0) Minimax Regret

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

View Full Document
Maximize Expected Value Papers Demanded Papers ordered 6 7 8 9 10 Expected Value 6 30 ¢ 30 ¢ 30 ¢ 30 ¢ 30 ¢ 30 ¢ 7 10 ¢ 35 ¢ 35 ¢ 35 ¢ 35 ¢ 30 ¢ 8 - 10 ¢ 15 ¢ 40 ¢ 40 ¢ 40 ¢ 25 ¢ 9 - 30 ¢ - 5 ¢ 20 ¢ 45 ¢ 45 ¢ 15 ¢ 10 - 50 ¢ - 25 ¢ 0 ¢ 25 ¢ 50 ¢ 0 ¢ Max Expected Value
Utility Theory n Consider a lottery where with probability pi you win ri n Tree representation of lottery (1/4, \$500; ¾, \$0): 1/4 3/4 \$500 \$0

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

View Full Document
n Suppose you had to choose between two lotteries L1 and L2: n L1 n Would you prefer L1, with expected value \$10K, or L2, with expected value \$15K? \$10,000
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 10/27/2010 for the course ESI 6321 taught by Professor Josephgeunes during the Spring '07 term at University of Florida.

### Page1 / 31

APME_6 - Applied Probability Methods for Engineers Slide...

This preview shows document pages 1 - 11. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online