pp

wafflejourneyΤεχνίτη Νοημοσύνη και Ρομποτική

14 Νοε 2013 (πριν από 3 χρόνια και 8 μήνες)

96 εμφανίσεις

Information Systems
to Support Decision
-
Making

BSAD 141

Dave Novak


BDIS: 2.1 (40
-
61)

Questions?

Topics Covered


Decision
-
making and types of decisions


Decision
-
making and the organizational
pyramid


Types of Decision Support Systems


Metrics


CSFs and KPIs


How they are related and different


Enhancing DSS with MIS


The Decision
-
Making
Process


The six
-
step decision
-
making process

Decision
-
Making Essentials

Organizational Pyramid

Decision
-
making and
problem
-
solving
occur at each level in
an organization



Decision
-
Making
Essentials

Types of Decisions



S
tructured





Unstructured


Decision
-
Making
Essentials

Information Requirements

Source:
Gelinas
, Sutton, and
Fedorowicz
, 2004

Decision
-
Making Essentials


Operational decision
making

OPERATIONAL

Decision
-
Making Essentials


Tactical / Managerial decision
making

MANAGERIAL

Decision
-
Making Essentials


Strategic decision making



STRATEGIC

Why Does This Matter?


The decision
-
makers at
different
managerial levels
are responsible for
making
very

different types of
decisions


some decisions are much more structured
than others


Each decision
-
making level requires
very

different types of information


MIS are more useful with respect to some
decisions



Supporting Decision
-
Making with MIS

Decision
-
Making Level from Organizational Pyramid

Generalized types of
Support Systems


Operational
-
level Support Systems



Managerial/Tactical
-
level Support
Systems



Strategic
-
level Support Systems



Operational Support
Systems


Transaction processing system (TPS)


Basic business system that serves the
operational level and assists in making
structured decisions


Online transaction processing (OLTP)

-

Capturing of transaction and event
information using technology to process,
store, and update


Source document


The original
transaction record


Operational Support Systems

Systems Thinking View of a TPS

Managerial Support
Systems


Online analytical processing (OLAP)



Manipulation of information to create
business intelligence

to support
decision making


Typically involves multidimensional
data and aspects of data mining


Complex analytical queries


Managerial Support
Systems


Decision support system (DSS)


Computer
-
based model(s) to
support managers and business
professionals during the decision
-
making process


Includes a combination of data and
analytical tools



Decision
Support Systems


Four types of quantitative models used by DSSs
include


1) What
-
if analysis



Decision
Support Systems


Four types of quantitative models used by DSSs
include


2) Sensitivity analysis

Decision
Support Systems


Four types of quantitative models used by DSSs
include


3) Goal
-
seeking analysis



Decision
Support Systems


Four types of quantitative models used by DSSs
include


4) Optimization analysis

Decision
Support Systems


Examples of decision support
modeling problems I have worked on


Network bandwidth allocation


New transportation project
prioritization


Identifying the most “critical” links in a
road or communications network


Blood bank


RBC ordering by type


Accessibility to emergency services

Decision
Support Systems

Systems Thinking View of a DSS

Managerial Support
Systems

Interaction Between a TPS and DSS

Strategic Support
Systems

Information becomes less granular as one moves up the
decision
-
making pyramid

Strategic Support Systems


Executive information system (EIS)



A
specialized DSS that
supports senior level
executives

within the organization


Focuses on summary / high
-
level information


Granularity


Visualization



Digital dashboard


Strategic Support Systems

Interaction Between a TPS and EIS

Strategic Support Systems



Most EISs offering the following capabilities


Ability to consolidate



Ability to drill
-
down




Ability to slice
-
and
-
dice



Metrics: Measuring
Success / Performance


Metrics




Metrics: Measuring
Success / Performance


Critical success factors (CSFs)


The crucial
steps companies make to perform to achieve their
goals and objectives and implement strategies


Create high
-
quality products


Retain competitive advantages


Reduce product costs


Increase customer satisfaction


Hire and retain the best professionals


Critical Success Factors
(CSFs)


CSFs are larger picture performance
measures
that directly relate to a specific goal



When you look at the individual CSF examples,
you should note that they don’t refer to specific
quantifiable measurements but provide specific
focus areas for achieving the goal


Metrics: Measuring
Success / Performance


Key performance indicators (KPIs)


The
quantifiable metrics a company uses to evaluate
progress toward critical success factors


Turnover rates of employees


Number of product returns


Number of new customers


Average customer spending


Key Performance Indicators
(KPIs)


Much like objectives are tied to goals, multiple
KPIs can be tied to a specific CPI




When you look at the individual KPI examples,
you should note that they CAN be explicitly
measured


Metrics: Measuring
Success / Performance


External KPI


Market share


The portion of the market
that a firm captures (external)



Internal KPI


Return on investment (ROI)


Indicates
the earning power of a project

http://
www.youtube.com/watch?v=R0rdMMfFyPQ


Efficiency Versus
Effectiveness Measures



MIS
Efficiency

metrics


The extent to which a firm is
using its
resources in an optimal way



getting
the most from its resources


Measure the performance of MIS itself,
such as throughput, transaction speed,
and system availability



Efficiency Versus
Effectiveness Measures



MIS
Effectiveness

metrics


How
well a firm is achieving its goals
and objectives


Measures the impact MIS has on
business processes and activities,
including customer satisfaction and
customer conversation rates


EFFICIENCY


-
VS
-

EFFECTIVENESS

Throughput



the amount of
information that can travel through
a system



Usability



the ease with which
people perform transactions or find
Info

Transaction speed



the amount
of time a system takes to perform
a transaction



Customer satisfaction



measured
by satisfaction surveys, how many
retained, and increase in revenue per
customer

System availability



the number
of hours a system is available



Conversion rates



how many
‘touches’ it takes to convert a first time
畳u爠r漠o散e浥m愠捵獴潭s爠慮搠
灵牣桡獥s瑨t⁰牯摵捴

Information accuracy



How
often a system generates the
correct results when doing the
same transaction many times



Financial



ROI, cost
-
benefit
analysis, break
-
even analysis

Response time


how long it
takes to respond to user
interactions.





Relationships Between Measures


Benchmark





Benchmarking






You can’t Improve what you don’t measure!!


Artificial Intelligence (AI)


In general, involves the study of man
-
made computer/technological systems
that exhibit some form of “intelligence”


How do you determine whether
something is “intelligent”?


The
Turing test
?



Common sense?


Artificial Intelligence (AI)


It is reasonable to say that, so far,
computers cannot replace people with
respect to making complex and
subjective judgment calls


Computers can be trained to mimic
human thinking and decision
-
making
though…


Artificial Intelligence (AI)


Five common categories of AI

1. Expert system

2. Neural Network

3. Genetic Algorithm

4. Computer
-
Based Agent

5. Virtual Reality

1. Expert System


A computerized system that is designed to
mimic the decision
-
making of a human
“expert” in a particular problem domain


Often employs if
-
then rules for decision
-
making


Can’t address unusual events or events not
considered during development


Often employs basic “true
-
false” logic


1. Expert System


Three parts of an
Expert System

User

User

Interface

Inference

Engine

Knowledge

Base

1)
User Interface


2)
Inference Engine


3)
Knowledge Base

Source
: Griffin and Lewis http://www.cs.uky.edu/~lewis/papers/inf
-
engine.pdf

1. Expert System


Three parts of an
Expert System


Inference Engine


procedural “rules” that
utilize user input and then execute actions
based on specific conditions


Knowledge Base


encoded representation
of domain expertise


Interface


dialog between the user and the
system is conducted via the user interface

1. Expert System


IF
-
THEN rules


IF (has 4 legs)


AND IF (has tail)


AND IF (is mammal)


AND IF (meows)



THEN it
is a
cat



An example of an expert system is medical
diagnostic software


Antecedents

Conclusion

1. Expert System


If you are interested, go through the
inference tutorial


http://www.expertise2go.com/webesie/tutoria
ls/Inference/


Or view
Negnevitsky

(2002)


ftp
://
ftp.dca.fee.unicamp.br/pub/docs/vonzub
en/ea072_2s06/notas_de_aula/Lecture02.p
df







2. Neural Networks


A computerized model that emulates
the functionality of the human brain


Often based on pattern recognition


The NN “learns” through trial and
error using a basic set of examples to
“self program” over time


Fuzzy logic



A mathematical
method of handling imprecise or
subjective
information

2. Neural Networks


Interested students can refer to


http
://pages.cs.wisc.edu/~
bolo/shipya
rd/neural/local.html


http
://www.willamette.edu/~
gorr/class
es/cs449/intro.html


A number of real
-
world applications
involve autonomous robotic
movement


military, space,
manufacturing

3. Genetic Algorithm


An adaptive
heuristic search
algorithm or model that solves
problems by mimicking an
evolutionary, survival
-
of
-
the
-
fittest
approach


A “fast” and suitable solution method
that is not guaranteed to be optimal


Interested students can refer to:
http
://
www.ai
-
junkie.com/ga/intro/gat1.html



3. Genetic Algorithm


Real
-
world examples include evolving
tactical plans for the military and route
choice and scheduling problems




4. Computer
-
Based Agent


A computer
-
based agent or intelligent
agent is a program that performs a service
on a regular schedule without user
intervention


An agent operates by sensing its
environment, gathering information, and
then acting on that information via some
set of instructions



4. Computer
-
Based Agent


Simplistic agent
-
based software is
particularly useful for repetitive tasks and
reminders


An
example of a simplistic computer
-
based
agent program is one that performs file
backup at a particular time


Other agent programs might perform a
new
search for a particular product that is
triggered whenever a
website
is updated

4. Computer
-
Based Agent


Interested students can refer to
Russel

and
Norvig

(1995
), Chapter 2
http://www.cs.berkeley.edu/~
russell/aima1e/
chapter02.pdf


And/or Tran and
Tran
http://
groups.engin.umd.umich.edu/CIS/cou
rse.des/cis479/projects/agent/Intelligent_ag
ent.html


5. Virtual Reality


“Almost” or “nearly” reality based on what
we (humans) experience


A 3
-
dimensional, computer
-
generated
environment that can be explored by a
person


Typically involves a head
-
mounted display
or glasses and may include additional
sensory stimuli such as sound and “touch
-
like” sensation

5. Virtual Reality


Interested students can
refer to


http
://
www.teach
-
ict.com/technology_explained/virtual_reality
/virtual_reality.html


http://
electronics.howstuffworks.com/gadget
s/other
-
gadgets/virtual
-
reality.htm



Examples include flight simulators

Artificial Intelligence (AI)


Robotics video
http
://
science.discovery.com/tv
-
shows/brink/videos/brink
-
artificial
-
intelligence
-
arrives.htm


Lecture Summary


6 steps of decision making


2 general types
of decisions


Different decision
-
making levels and
the
organizational pyramid


The different types
of Decision Support
Systems


Metrics


CSFs and KPIs


How they are related and different


Artificial intelligence categories