Analytical Techniques
Singapore International Executive MBA
Rutgers University
July/August 2017
Professor Jonathan Eckstein
Last update time:
Sunday, August 06, 2017 03:35:22 -0400
General Information and Links
Class 1 (Saturday) -- Introduction; Linear Optimization Models
Class 2 (Sunday) -- Integer and Nonlinear Optimization Models
- Solutions to the paper recycling and gas trading problem assigned in class
during the last module on Saturday are posted on BlackBoard
(under "Course Documents", then "Class of Sunday, July 30")
- Introduction: Employee scheduling model
- Binary variables: capital budgeting model
- Assignment grids: Machinco spreadsheet
model
- Milkem spreadsheet model
- Remington fixed charge problem
- Allocated cost spreadsheet example
- Simple nonlinear model: EOQ spreadsheet
- Single product pricing
spreadsheet
- Solutions to in-class problems will be posted on BlackBoard
(under "Course Documents", then "Class of Sunday, July 31")
- Assignment due in class on Tuesday, August 1:
- Consider the situation in Case 6.4 on pages 311-312 of the textbook.
- You do not need to answer the question 4 or 5 in the textbook
(for your own amusement, however, you may wish to think about question 5; I will show a
solution in class)
- Hand in an algebraic formulation of the problem; assume, as
mentioned in question 1 in the textbook, that it costs CA$0.10 to
transport one cubic meter of snow one kilometer.
- Solve the problem using Excel and Solver (which effectively answers
questions 1-3 in the text).
- Also hand in your spreadsheet solution, carefully following the instructions on
page 11 of the course pack. The result should look like the spreadsheet
exhibits in the course pack: one sheet showing the solution, and then a second
version displaying all formulas, with annotations indicating the decision
variables, the objective function, and all the constraints (but your annotations
can be handwritten if you like). I have limited time to grade the
assignments; if you do not follow the instructions, grading will take too long,
and I may deduct points.
- Please hand in "hard copy" (paper); electronic submission is allowed only if you
are traveling and have made prior arrangements with me (electronic submissions take longer to grade).
- An Excel file containing the raw data for
the problem is available.
Class 3 (Tuesday Night) -- Nonlinear and Multi-Objective Optimization
Models, Introduction to Probabilistic Modeling
- Solutions and other material relating to the homework case are posted on BlackBoard
under the
"course documents" section, folder "Homework
1". The files will become accessible after class begins.
- See "Repaired for Take-Off", pages 31-36 of the course pack,
for an example of a decision support system (DSS)
- Here
is a more recent example of a successful decision support system, from UPS
- More realistic pricing
problem (MULTIPRICING) spreadsheet
- Multiple objective
example: Blackstone Mining
- Introduction to uncertain decision making and expected monetary value: hotel site spreadsheet
- Practice material for the first exam, including solutions, will be posted on BlackBoard
under "course documents" and then "practice material for first
exam"). The exam format will be one
algebra formulation question and one spreadsheet question, in respectively
similar formats to the practice questions on BlackBoard (but I will try to
make the algebra question relatively easy). You may refer to your
textbook and personal notes during the exam. You may not use your
computer during the exam -- if you took notes on your computer, please
print them out before the exam.
Class 4 (Thursday Night) -- First Exam, Decision Trees
- The first half of the class will be a 90-minute exam
- See the practice material on BlackBoard
- There will be a spreadsheet-format (but answered on paper) question in a
similar format to those in the practice material.
- There will be one algebra-format question in a similar format to those
in the practice material
- Decision trees (these are created with TreePlan,
but they can easily be drawn by hand)
- Homework case assignment for Saturday
Class 5 (Saturday) -- Bayes' Formula and Introduction to Simulation
- The Eagle Credit Union problem (problem 14.25, page 802 of the textbook) is analyzed on pages
63-65 of the
coursepack
-
Download the YASAI add-in (just save and do not open directly;
then follow installation directions in the YASAI User Guide)
- Our first simulation example: the classic newspaper stocking problem
-
Piedmont Airlines example
- Binomial random variables are described on page 74 of the coursepack
- Hiring translators in-class exercise
- Poisson distributions:
- The idea of Poisson distributions is described on pages 74-77 of the
coursepack
- Graphs demonstrating
Poisson distribution as a limit of
binomial distributions
- Another airline overbooking example
model, but with Poisson demands and two classes of service (passengers
not seated formulas simplified)
- Power generation example
demonstrating simulation with continuous random variables
- The solution to the chemical simulation and diesel fuel prepositioning
problems are on
BlackBoard
- The assignment for the last class is to read the North Star case in the back of
the course pack (it would take much too long to read in class)
- You should also study for the second exam. Practice material is available on
BlackBoard.
Class 6 (Sunday) -- More Simulation, Second Exam
- Practice material for the second test is posted on BlackBoard
- Insurance reserve capital
model (a model with continuous random variables)
- The second case study homework solution is posted on BlackBoard
- An analysis of the North Star case (both by decision tree and simulation) is
posted on BlackBoard
- Multi-period simulation models:
- Discrete-event simulation software (Arena,
in this case) was shown during lunch
- Exam review problem: a decision problem tree from a past exam
(questions and solutions on BlackBoard)
- Second exam
- DONE!
