2. DSS
modeling
–
Issues
• DSS
–
can
be
composed
of
mul8ple
models
• Modeling
Issues
-‐
• Iden8fica8on
of
problems
and
environment
analysis
• Variable
iden8fica8on
• Forecas8ng
(predic8ve
analysis)
3. DSS
modeling
–
Categories
• Op8misa8on
of
problems
with
few
alterna8ves
• Op8misa8on
via
algorithm
• Op8misa8on
via
analy8cal
formula
• Simula8on
• Heuris8cs
• Predic8ve
models
• Other
Models
5. DSS
modeling
–
Trends
• Model
libraries
and
solu8on
techniques
• Using
web
tools
–
perform
modeling,
op8misa8on,
simula8on
etc
• Mul8dimensional
analysis
• Model
for
model
analysis
6. Classifica8on
of
DSS
Models
Sta$c
Analysis:
• Sta8c
model
takes
a
single
snapshot
of
situa8on
• Everything
occurs
in
a
single
interval.
• E.g.
Make
or
buy
decision
• Stability
of
the
relevant
data
is
assumed.
7. Dynamic
Analysis:
• Represents
scenarios
that
change
over
8me.
• E.g.
5-‐year
profit
and
loss
projec8on
in
which
the
input
data,
such
as
costs,
prices,
and
quan88es,
change
from
year
to
year.
• Time
dependent
• Important
because
they
use,
represent,
or
generate
trends
and
paRerns
over
8me.
• Shows
average
per
period,
moving
averages
and
compara8ve
analysis.
8. Certainty,
uncertainty,
and
risk
Decision
situa8ons
are
oSen
classified
on
the
basis
of
what
the
decision
maker
believes
about
the
forecasted
results.
The
categories
are:
• Certainty
• Risk
• Uncertainty
9. Decision
Making
Under
Certainty
• Complete
knowledge
is
available
• Decision
maker
knows
the
outcome
of
each
course
of
ac8on
• Situa8on
involve
is
oSen
with
structured
problems
with
short
8me
horizons
• Certain
models
are
rela8vely
easy
to
develop
and
solve
and
they
can
yield
op8mal
solu8ons.
10. Decision
making
under
uncertainty
• Several
outcomes
for
each
course
of
ac8on.
• Decision
maker
does
not
know,
or
cannot
es8mate
the
possible
outcomes.
• More
difficult
because
of
insufficient
informa8on.
• Involves
assessment
of
the
decision
maker’s
aXtude
towards
risk.
11. Decision
making
under
risk
(Risk
analysis)
• Decision
maker
must
consider
several
possible
outcomes
for
each
alterna8ve.
• The
decision
maker
can
assess
the
degree
of
risk
associated
with
each
alterna8ve.
• Risk
analysis
can
be
performed
by
calcula8ng
the
expected
value
for
each
alterna8ve
and
selec8ng
the
one
with
best
expected
value.
12. Decision
analysis
with
decision
tables
and
decision
trees
Decision
Table:
• Organize
informa8on
and
knowledge
in
systema8c
tabular
manner
13. Decision
Trees:
• Alterna8ve
representa8on
of
the
decision
table
• Shows
the
rela8onship
of
the
problem
graphically
and
handle
complex
situa8ons
• Can
be
cumbersome
if
there
are
many
alterna8ves
or
sta8c
nature.
• TreeAge
Pro
and
Precision
Tree:
Powerful
and
sophis8cated
decision
tree
analysis
systems
14. Structure
of
mathema8cal
models
for
decision
support
Components
of
decision
support
mathema8cal
models:
• Result
Variables
• Decision
Variables
• Uncontrollable
variables
• Intermediate
result
variables
15. • Result
Variables:
reflect
the
level
of
effec8veness
of
a
system
• Decision
Variables:
describes
alterna8ve
course
of
ac8on.
• Uncontrollable
Variables:
Some
factors
that
affect
the
result
variables
but
not
under
the
control
of
decision
maker.
• Intermediate
result
Variables:
reflect
intermediate
outcomes
in
mathema8cal
models.
17. Sensi8vity
Analysis
• ARempts
to
assess
the
impact
of
a
change
in
input
data
on
proposed
solu8on.
• Important
because
it
allows
flexibility
and
adapta8on
to
changing
condi8ons
• Provides
a
beRer
understanding
of
the
model
and
the
decision
making
situa8on
• Used
for:
1.Revising
models
to
eliminate
too-‐large
sensi8vi8es.
2.Adding
details
about
sensi8ve
variables.
3.Obtainong
beRer
es8mate
of
sensi8ve
external
variables.
4.Altering
a
real-‐world
system
to
reduce
actual
sensi8vi8es.
18. What-‐If-‐Analysis
• What
will
happen
to
the
solu8on
if
an
input
variables,
an
assump8on,
or
a
parameter
value
is
changed
• With
the
appropriate
user
interface,
it
is
easy
for
manager
to
ask
a
computer
model
different
ques8ons
and
get
the
answers.
• Common
in
expert
systems.
• User
get
an
opportunity
to
change
their
answers
to
some
ques8on’s.
19. Goal
Analysis
• Calculates
the
values
of
the
inputs
necessary
to
achieve
a
desired
level
of
output.
• Represents
a
backward
solu8on
approach
20. Problem
solving
search
methods
The
choice
phase
of
problem
solving
involves
a
search
for
an
appropriate
course
of
ac8on.
Search
approaches
are:
• Analy8cal
Techniques
• Algorithms
• Blind
Searching
• Heuris8c
Searching
21. Simula8on
• Is
a
appearance
of
reality.
• A
technique
for
conduc8ng
experiments
with
computer
on
model
of
a
management
system
• Characteris$cs:
1.Simula8on
typically
imita8ve.
2.Technique
for
conduc8ng
experiments.
3.Descrip8ve
rather
than
a
norma8ve.
4.Used
only
when
a
problem
is
too
complex
to
be
treated
using
numerical
op8mizing
techniques.
22. Advantages
of
simula8on
• Theory
is
fairly
straighcorward.
• Great
8me
compression
• Descrip8ve
rather
than
norma8ve.
• Built
from
the
manager’s
perspec8ve.
• Built
for
one
par8cular
problem
and
cannot
solve
any
other
problem.
•
A
manager
can
experiment
to
determine
which
decision
variables
and
which
part
of
environment
are
really
important,
and
with
different
alterna8ves.
23. • Can
handle
an
extremely
wide
variety
of
problem
types,
such
as
inventory
and
staffing.
• Can
include
the
real
complexi8es
of
problems.
• Automa8cally
produce
many
important
performance
measures.
• Rela8vely
easy-‐to-‐use
simula8on
packages.
• OSen
the
only
DSS
modeling
method
that
can
readily
handle
rela8vely
unstructured
problem.
24. Disadvantages
of
simula8on
• An
op8mal
solu8on
cannot
be
guaranteed.
• Model
construc8on
can
be
a
slow
and
costly
process.
• Solu8ons
are
not
transferable
to
other
problems
• Easy
to
explain
to
managers
that
analy8c
methods
are
overlooked.
• Requires
special
skills
because
of
the
complexity
of
the
formal
solu8on
method.
25. The
Methodology
of
Simula8on
Test
&
validate
the
model
Real
world
problem
Define
the
problem
Construct
simula8on
model
Implement
the
result
Design
the
simula8on
experiments
Conduct
the
experiments
Evaluates
the
results
26. Simula8on
type
Probabilis8c
Simula8on:
• One
or
more
of
the
independent
variables
• Follow
certain
probability
distribu8ons
namely
1.Discete
distribu8on
2.Con8nuous
distribu8on
• Conducted
with
the
aid
of
technique
called
Monte
Carlo
simula8on.
27. Time-‐Dependent
Vs
Time-‐Independent
Simula8on:
• Time-‐independent-‐not
important
to
know
the
exact
8me
of
event
• Time-‐dependent-‐In
wai8ng
line
problems,
it
is
important
to
know
the
precise
8me
of
arrival.
28. Object-‐Oriented
Simula8on:
• SIMPROCESS
is
an
object-‐oriented
process
modeling
tool
that
allows
user
to
create
a
simula8on
model
by
using
screen
based
object.
• Unified
Modeling
Language(UML)-‐
Designed
for
object-‐oriented
and
object
based
systems
and
applica8ons.
• Java
based
simula8ons
are
essen8ally
object
oriented.
29. Visual
Simula8on:
• Graphical
display
of
computerized
results
• Includes
anima8ons
• Is
one
of
the
most
successful
development
in
computer-‐human
interac8ons
and
problem
solving.
30. Quan8ta8ve
SoSware
Packages
• Are
preprogrammed
models
and
op8miza8on
systems.
• Serve
as
building
blocks
for
other
quan8ta8ve
models
• A
variety
of
these
are
available
for
inclusion
in
DSS
as
major
and
minor
modeling
components.
• Revenue
management
systems
focus
on
iden8fying
right
product
for
right
customer.
• Airlines
have
used
such
systems
to
determine
right
price
for
each
airline
seat.
• System
also
available
for
retail
opera8ons,
entertainment
venues,
and
many
other
industries.