Problem Solving and Decision Making
Nov 27, 2014
Problem Solving and Decision Making
Decision TheoryEven SwapsStatistical Quality Control for Process
DevelopmentDecision-Making MethodsDecision AnalysisData-based Decision Making and CRM
Presentation Overview
Decision Theory
Decision Theory can be used to determine optimal strategies when a decision maker is faced with several decision alternatives and an uncertain or risk-filled pattern of future events
e.g. Production of a product before the demand is actually known
Decision Theory
Pay-off Tables
The future events which are not under the control of decision maker, are referred to as “states of nature”
Pay-off Tables contains any measure of output that may be appropriate for a particular situation e.g cost, profits etc
High Acceptance Low AcceptanceDecision Alternatives s1 s2Lease a large system d1 200,000 (20,000) Lease a medium-sized system d2 150,000 20,000 Lease a small system d3 100,000 60,000
States of Nature
Profit or payoff in $
Types of decision making situations Decision making under certainty:
The process of choosing a decision alternative when the state of nature is known only one column in the payoff table optimal decision is the one corresponding to the best pay off in the column
Decision making under uncertainty: The process of choosing a decision alternative when the state of
nature is not known
Decision Making Situations
Decision making under Uncertainty
Pessimistic approach Maximize the minimum possible payoff (profit
here) Concentrates on worst possible payoffs
Maximin Decision Criterion
High Acceptance Low AcceptanceDecision Alternatives s1 s2Lease a large system d1 200,000 (20,000) Lease a medium-sized system d2 150,000 20,000 Lease a small system d3 100,000 60,000
States of NatureMaximum
of the minimum
payoff values
Optimistic approach Maximize the maximum possible payoff (profit
here) Concentrates on best possible payoffs
Maximax Decision Criterion
High Acceptance Low AcceptanceDecision Alternatives s1 s2Lease a large system d1 200,000 (20,000) Lease a medium-sized system d2 150,000 20,000 Lease a small system d3 100,000 60,000
States of Nature
Maximum of the
maximum payoff values
Compute the expected value for each decision alternatives
Select the alternative yielding the best expected value
Expected Monetary Value
EMV(di )= ∑ P(Sj) V(di, Sj) j= I to N
Even Swaps
A Rational Method for making Trade-Offs A practical way of making trade-offs among any
set of objectives across a range of alternatives
Steps Involved: Creating a consequences Table Eliminating “Dominated” Alternatives Making Even Swaps
Even Swaps
Even Swaps
Statistical Quality Control for Process Development
An approach to solving problems in Production Processes
Can also be used to determine whether the quality of an output is “acceptable” or not.
Designed to find underlying problems which cause defects or other undesirables results, and fix them permanently
Statistical Quality Control for Process Development
Statistical Quality Control for Process Development
Pareto Chart
Contains both bars and a line graph, where individual values are represented in descending order by bars, and the cumulative total is represented by the line.
Fishbone diagrams, cause-and-effect diagrams
A method of showing possible causes of a problem.
An aid to brainstorming and hypothesis generation
Decision-Making Methods
Consensus Method is a group decision making process that seeks not only the agreement of most participants but also the resolution or mitigation of minority objections.
Dialectical Inquiry First group develops a detailed set of recommendations. The 2nd subgroup then develops a reasonable set of opposing recommendations. After discussions, both groups agree on a final, common list of assumptions and recommendations.
Devil’s Advocacy First group develops a detailed set of recommendations. The 2nd subgroup then develops a reasonable set of plausible critique to the other subgroup. The first group then revises its initial proposal and continues until both subgroups can agree.
Decision-Making Methods
Decision Analysis
A decision tree is a decision support tool that uses a tree-like graph or model of decisions and their possible consequences, including chance event outcomes, resource costs, and utility.
It is one way to display an algorithm.
Decision Trees
Decision Trees
Data-based Decision Making and CRM
Case: TESCO
SegmentationSegmentation
Criteria for need-based segments:adequate size cluster principle: members similar and distinct from
members of other segments segments must be addressable and profitable
Definition of Segment A group of customers who share the
same or similar needs
Tesco and CRMTesco and CRM
Why has TESCO been successful?
Lifestyles are built from a customer’s “DNA” profile at Tesco: Miss Jones’ DNA profile
• 25,000 products are used to create a customer’ DNA
• Tesco has a DNA profile for every single customer
A “Time Poor Food Rich” customer
Source: © Dunnhumby, 2006
-2
-1
0
1
2
3
4
Adven
turo
us
Big B
ox
Brand
ed
Calo
rie
Coun
ters
Calo
rie
Load
ers
Conv
enie
nce
Fine
st
Fore
ign
Fres
h
Gre
en
Hea
lthy
Hig
h Pr
ice
Kids
Long
Life
Loos
e
Low P
rice
Lunc
h Box
Mea
t
Pre
Pack
ed
Cook
ing
from
Scr
atch
Smal
l Box
Tesc
o's Own
Trad
itio
nal
Value
Veget
aria
n
TelCo in E. Europe: TelCo in E. Europe: The Consumer SegmentsThe Consumer Segments
AzerbaijanAzerbaijan GeorgiaGeorgia KazakhstanKazakhstan MoldovaMoldova
6%13%
25%
12%24%
20%
6%16%11%
15%
26% 26%
10%22%
17%
9%21%
22%
16%24%
5%
8%
20%28%
Talk n’ Texters(TT)
Talkative Trendies (TT)
Wannabes(WB)
Sophisticated Careerists(SC)
Gaming Youths(GY)
Laggards(L)
Thank You!