Negotiation
Domestic Focus
Class 4
Prof. Silvia Saccardo
Fall 2016
Present Bias: tendency to over-value the present at the
expense of our long-term intentions
Present Bias: tendency to over-value the present at the
expense
Negotiation
Domestic Focus
Class 2
Prof. Silvia Saccardo
Fall 2016
Some admin info
Class is on blackboard
My email: ssaccard@andrew.cmu.edu
Office hour: Wed 3-4pm
TA: Ammara Shariq
Email: ashariq@andrew.cmu.edu)
Office hour: Mond
Negotiation
Domestic Focus
Class 3
Prof. Silvia Saccardo
Fall 2016
Questions
List a behavioral principle for why first offers
Questions
List a behavioral principle for why first offers
Discuss with person next to you.
Qu
DADSS
Optimization:
Genetic Algorithms for Rugged Landscapes
Part II
1
Navigating Rugged Landscapes
We can run genetic algorithms in Excel using
Evolutionary Solver
Lets try it for the Campaigning President problem
2
Campaigning President Model
Input
Repr
HOMEWORK ASSIGNMENT #10
Due:
Thursday, April 28, 2016 (beginning of class)
1. Choose a team of up to 4 applicants to hire for a special project. The team must satisfy the following
requirements:
There must be at least 1 male and 1 female
There should be
Lecture 26
DADSS
Linear Programming, Part III
Last updated: 4/19/2016
1
Administrative
Homework #9 due Thursday
2
Personal Training
Inspired by recent media about boxers, you have decided to build up muscle and
begin training (so that you can be hired t
USE YOUR SMC NUMBER INSTEAD OF YOUR NAME
ON ALL ASSIGNMENTS
HOMEWORK ASSIGNMENT #1
Due: Tuesday, January 19, 2016 (beginning of class)
Find an article in a newspaper that describes a person or organization making a decision.
Using the techniques discussed
DADSS
Optimization:
Genetic Algorithms for Rugged Landscapes
Part I
1
Optimization
How do we navigate rugged landscapes?
A rugged
Landscape
2
Cant We Just Search Through
Every Possible Solution?
We can program a computer to search through every
possible s
Lecture 25
DADSS
Linear Programming, Part II
Last updated: 1/6/2016
1
Administrative
Homework #9 due next Thursday
2
Taylors Toy Cars and Trucks (V2)
Taylor mass produces and sells toy cars and trucks and wants to
maximize profits for the year. Taylor s
HOMEWORK ASSIGNMENT #6
Due: Tuesday, March 1, 2016 (beginning of class)
Recruit two of your friends or classmates to answer the following questions. Alternate
whether you ask A or B first. Always ask C last.
Consider an opportunity to extend your life in
HOMEWORK ASSIGNMENT #9
Due: Thursday, April 21, 2016 (beginning of class)
1) Your company makes two types of mobile homes: the Enviro and the Rustic. The expected
profit for the Enviro is $8,000 per unit, while for the Rustic, it is $5,000 per unit. Maxim
Lecture 24
DADSS
Introduction to Linear Programming
Last updated: 4/9/2015
1
The (Simple) Hiring Problem
that youre trying to hire or admit people and you have 5000 applicants for
Suppose
1 position.
Youve arranged the files by GPA and years of experience
Lecture 19
DADSS
Evaluating Simulation Output
Last updated: 12/9/2015
1
How many simulations must we run?
With many simulations, we converge on the right
answer, but we might never get an exact solution
When do we get close enough?
2
How many simulations
DADSS
Lecture 23
Optimization:
Simple and Rugged Landscapes
What do these decision have
in common?
How much money should the government
invest in green energy R&D grants and how
should it be allocated?
How much should Apple charge for an app?
Which fli
Lecture 18
DADSS
Simulation
Last updated: 12/9/2015
1
Staffing a 911 call center
How many people do we need to staff at a 911 call
center?
2
Staffing a 911 call center
How many people do we need to staff at a 911 call center?
Possible objectives:
Minimi
Lecture 20
DADSS
Modeling a Queue
Last updated: 12/9/2015
1
More randomness
Much of our simulations have been based on uniformly
distributed random numbers
Dice rolls, coin flips,
Many real-world questions depend on non-uniform randomness
Car-acciden
DADSS Queuing Handout
M/M/1 Queue
Assume that both the arrival and service processes are exponential/Poisson and that
there is a single server.
Exponential distribution for interarrival times using arrival rate
CDF : F(x) 1-e x
Inverse CDF : x
1
ln(1 F
HOMEWORK ASSIGNMENT #5
Due: Tuesday, February 23, 2016 in Class
(1) Suppose that you work for the snow removal program for the City of Pittsburgh.
They are trying to decide what supplies to order for next year. Ice control COSTS
depend on the temperature
HOMEWORK ASSIGNMENT #2
Due: Thursday, January 28, 2016 (beginning of class)
1) Sylvie is choosing an apartment to move into from the five alternatives listed below.
Sylvie tells you she considers the following exchanges to be even swaps:
$20 per month to
Lecture 12
DADSS
Decision Analysis
Expected Value of Imperfect Information
1
Imperfect Information
Last class, we discussed the basic
model for using Bayes Theorem to
solve decision problems under
imperfect information
Today:
Calculating EVII
Multiple
Lecture 10
Expected Value of Perfect
Information
EVPI
1
Athabasca oil sands
Alberta, Canada has a large amount of petroleum in what is known as oil
sands (or tar sands)
Keystone XL pipeline was proposed to import oil from these oil sands into the US
Pi
Lecture 11
DADSS
Decision Analysis
Bayes Theorem
1
Information and Value
We have seen that
Information (often) has value
How much value information has depends on:
Whether we change our decisions
What we know at the outset (Prior or Initial Beliefs)
Th
PracticeProblems
ACarBuyingProblem
Youaretryingtodecideifyoushouldbuyausedcar. YouhavelookedatConsumer
Reports,anditsaysthattheparticularmodelthatyouareinterestedinisveryreliable.
The only potential problem is with the transmission. On about 20% of the ca