DECISION SUPPORT AND ARTIFICIAL INTELLIGENCE

1
CHAPTER 5

• DECISION SUPPORT AND ARTIFICIAL INTELLIGENCE
• Brainpower for Your Business

2
Opening caseA DSS to save lives

• Nuclear radiation
• 1986 two explosions in Unit 4 of the
  Chernobyl Nuclear Power Plant
• 3.5 million people high incidence of thyroid
  cancer in children
• Virginia TEDSS ---gt the best evacuation
  routes notify
• use information.
• Not static, fluctuate DSSs task --- find the
  best solution quickly
• the information comes from .
• Process the info in its simulation models, TEDSS
  provides output on the factors.
• With the outputs,evaluate traffic management
  strategies one way operation,shoulder
  use,flashing signals

3
The Survival of Your Business Depends on
Effective Decision Making
Introduction
5-2

• IT can supplement human decision making
• IT can (sometimes) replace human decision making

4
YOUR FOCUS IN THIS CHAPTER
Introduction
5-3

• Types of Decisions
• The Decision-Making Process
• IT Brainpower
• How IT Brainpower Supports the Decision-Making
  Process

See Figure 5.1, page 175
5
TYPES OF DECISIONS YOU FACE
Types of Decisions
5-4

• Buy salt and choosing a job
• STRUCTURED DECISION - involves processing a
  certain kind of information in a specified way
  so that you can always get the right answer.
• Eg calculating gross pay for hourly worker
• NONSTRUCTURED DECISION - is one for which there
  may be several right answers and there is no
  precise way to get a right answer.No rules or
  criteria exist that guarantee you a good
  solution.
• Somewhere in between choosing the job
• 
  stock market investment analysis

See Figure 5.2, page 175
6
TYPES OF DECISIONS YOU FACE
Types of Decisions
5-5

• By frequency
• RECURRING DECISION - is one that happens
  repeatedly, and often periodically.you will use
  the same set of rules each time
• NONRECURRING(ad hoc) DECISION - is one that you
  make infrequently, perhaps only once. You may
  have different criteria for determine the best
  solution each time
• Eg a company merger

7
PHASES OF THE DECISION-MAKING PROCESS
The Decision-Making Process
5-6

• 1. INTELLIGENCE -Find what to fix
• 2. DESIGN - Find fixes
• 3. CHOICE - Pick a fix prescriptive phase
• 4. IMPLEMENTATION - Apply the fix

See Figure 5.3, page 177
8
TYPES OF IT BRAINPOWER
IT Brainpower
5-7

• Decision Support Systems (DSS)
• Group Decision Support Systems (GDSS)
• Geographic Information Systems (GIS)
• Artificial Intelligence Systems (AI)
• Expert Systems
• Neural Networks
• Genetic Algorithms
• Intelligent Agents

9
A DECISION SUPPORT SYSTEM (DSS)
DSS
5-8
is a highly flexible and interactive IT system
that is designed to support decision making when
the problem is not structured. A DSS
assists you in making a decision, as opposed to
making the decision for you. The alliance
between you and a DSS,See Figure 5.4, page 177
DSS s objective improve your effectiveness
10
A DSS Can Help With Analysis Tasks Like
DSS
5-9

• Deciding where to spend advertising dollars
• Analyzing sales trend
• Analyzing drug interactions
• Developing airline schedules
• Pricing products

11
A DSS Has Three Components
DSS
5-10

• DATA MANAGEMENT - to allow you to store and
  maintain information.
• Information comes from one or more of
  three sources
• MODEL MANAGEMENT - to allow you to store and
  access models.
• A model is a representation of some event
  ,fact or situation. Take various forms
  cardboard cutouts race gameflight simulators
• USER INTERFACE MANAGEMENT - to allow you to
  communicate with the DSS
• Combine your knowledge with the storage and
  processing capability of computer
• the best user interface use your
  terminology

See Figure 5.5, page 178
12
STEPS IN DEVELOPING A DSS
DSS
5-11

• 1.INTELLIGENCE
• Do you need a DSS?
• What do you want the DSS to do?-- divide and
  conquer
• What information and models do you need?
• 2.DESIGN
• Whats available to you commercially?
• What can you build?
• 3.CHOICE
• What will you buy/build?
• 4.IMPLEMENTATION
• Build or install DSS
• Learn, test, and evaluate the DSS

13
A GROUP DECISION SUPPORT SYSTEM (GDSS)
GDSS
5-12
is a type of decision support system that
facilitates the formulation of and solution to
problems by a team. A GDSS helps a team to
generate ideas, identify strengths and
weaknesses, choose an alternative, and reach a
consensus.
14
A GDSS Can Help With Team Tasks Like
GDSS
5-13

• Determining new product design
• Evaluating the impact of new competition
• Formulating a strategic plan

15
THE STEPS OF GROUP DECISION MAKING
GDSS
5-14

• 1.Brainstorming - generate ideas
• 2.Issue Categorization and Analysis - classify
  ideas
• 3.Ranking and Voting - prioritize alternatives,
  reach consensus

See Figure 5.9, page 184
16
PEOPLE IN A GDSS
GDSS
5-15

• Team People united by a common goal
• GDSS facilitator has two roles
• Nontechnical run the meeting
• Technical operate IT components

17
IT TOOLS IN A GDSS
GDSS
5-16

• Groupware - software than enables team work
• DSS Capabilities - software that enables team
  decision making
• Telecommunications - hardware and software that
  connect computers

18
MEETINGS WHEN AND WHERE YOU WANT TO BE
GDSS
5-17

• Same-Time (Synchronous) Meetings
• In the same room
• In the same city
• On the same planet
• Different-Time (Asynchronous) Meetings
• In the same city
• On the same planet

19
A GEOGRAPHIC INFORMATION SYSTEM (GIS)
GDSS
5-18
is a decision support system designed
specifically to work with spatial information. A
GIS represents information in overlapping layers,
each of which shows a different aspect of the
total picture.
20
A GIS Can Help With Spatial Analysis Tasks Like
GDSS
5-19

• Identifying the best site to locate a branch
  office
• Targeting pockets of potential customers
• Repositioning promotions and advertising

21
ARTIFICIAL INTELLIGENCE (AI)
AI
5-20
is the science of making machines imitate human
thinking and behavior. Expert System Marion
County Department of Public Welfare case of
child abuse standardizing 600 cases 80
caseworker turnover rate a robot equipped
with simulated human senses and the capability of
taking action on its own Piedmont Hospital
Pharmacy Dosage Dispenser
22
AI systems are used in

• Finance analysis
• hospital
• government agencies
• credit card companies
• insurance companies

23
TYPES OF AI USED IN BUSINESS
AI
5-21

• EXPERT SYSTEMS
• NEURAL NETWORKS
• GENETIC ALGORITHMS
• INTELLIGENT AGENTS

70 of the top 500 companies use AI as part of
decision support.
24
AN EXPERT SYSTEM
Expert System
5-22
is an artificial intelligence system that applies
reasoning capabilities to reach a conclusion. An
expert system captures expertise from a human
expert and applies it to a problem. Expert
systems are excellent for diagnostic and
prescriptive problems.
25
An Expert System Can Perform Diagnostic and
Prescriptive Tasks Like
Expert System
5-23

• Auditing and tax planning
• Diagnosing illnesses
• Managing forest resources
• the difference between DSS and ES
• See table 5.1

26
COMPONENTS OF AN EXPERT SYSTEM
Expert System
5-24

• Information
• People
• IT Components

How do these components fit together? See the
traffic light example on page 185.
27
INFORMATION TYPES IN AN EXPERT SYSTEM
Expert System
5-25

• Domain Expertise(the core) - the set of problem -
  solving steps.
• Whys Information - the information that
  explains the expert systems actions.
• Problem Facts - specific information concerning
  the current problem. Symptoms or assertions

28
PEOPLE IN AN EXPERT SYSTEM
Expert System
5-26

• Domain Expert - the person who knows how to solve
  the problem without the aid of IT.
• Knowledge Engineer - the person who builds the
  expert system(formulate the domain expertise into
  ES).
• Knowledge Worker (user)- the person who uses the
  expert system to solve a problem.

29
IT COMPONENTS IN AN EXPERT SYSTEM
Expert System
5-27

• KNOWLEDGE BASE - stores the domain expertise
  (rules).
• INFERENCE ENGINE - take your problem facts and
  searches the knowledge to reach a conclusion.

30
IT COMPONENTS IN AN EXPERT SYSTEM
Expert System
5-28

• KNOWLEDGE ACQUISITION - used by the knowledge
  engineer to build the expert system.
• USER INTERFACE - used to run a consultation.
• EXPLANATION MODULE - stores the why
  information.
• Distinction between the development and use
  of ES .See figure 5.16

31
TO ACQUIRE AN EXPERT SYSTEM
Expert System
5-29

• Buy an off-the-shelf expert system
• see table 5.2
• Use an expert system shell
• provide you IT components except knowledge
  base
• Develop an expert system from scratch
• Lisp, Prolog or C
• Consider two things 1) amenable to ES solution
  1) domain expert

32
WHAT EXPERT SYSTEMS CAN DO
Expert System
5-30

• Handle massive amounts of information
• Reduce errors
• Improve customer service

• Provide consistency in decision making
• Provide new information
• Reduce costs

Clear rules and procedure
33
WHAT EXPERT SYSTEMS CANT DO
Expert System
5-31

• Handle all types of domain expertise
• transferring domain expertise to the ES is
  sometimes difficult
• automating the reasoning process may be
  impossible
• Solve problems other than those for which they
  are designed
• Apply common sense or judgment to a problem

34
Ethical Questions in ES

• Will an ES act ethically?
• An ES will act the way it is programmed to
  act
• What kind of decisions should you let ES make
  for you?
• Should you accept the decision of an ES without
  question?

35
A NEURAL NETWORK
Neural Network
5-32
is an artificial intelligence system which is
capable of learning because its patterned after
the human brain.
A neural network simulates the human ability to
classify things based on the experience of seeing
many examples.
36
A Neural Network Can Perform Pattern Recognition
Tasks Like
Neural Network
5-33

• Distinguishing different chemical compounds
• Detecting anomalies in human tissue that may
  signify disease
• Reading handwriting
• Detecting fraud in credit card use

NN works best on pattern-recognition problems for
which a vast amount of historical information is
available.
37
A GENETIC ALGORITHM
Genetic Algorithm
5-34
is an artificial intelligence system that mimics
the evolutionary, survival-of-the-fittest
processes to generate increasingly better
solutions to a problem.
Genetic algorithms produce several generations of
solutions, choosing the best of the current set
for each new generation.
38
A Genetic Algorithm Can Generate Lots of
Solutions As In
Genetic Algorithm
5-35

• Generating human faces based on a few known
  features.
• Generating solutions to routing problems.
• Generating stock portfolios.

39
THE CONCEPTS OF EVOLUTION IN GENETIC ALGORITHMS
Genetic Algorithm
5-36

• SELECTION - or survival of the fittest. The key
  is to give preference to better outcomes.
• CROSSOVER - combining portions of good outcomes
  in the hope of creating an even better outcome.
• MUTATION - randomly trying combinations and
  evaluating the success (or failure) of the
  outcome.

40
AN INTELLIGENT AGENT
Intelligent Agent
5-37
is an artificial intelligence system that can
move around your computer or network performing
repetitive tasks independently, adapting itself
to your preferences. An intelligent agent is
like a travel agent in that it performs tasks
that you stipulate.
41
An Intelligent Agent Can Perform Tasks Like
Intelligent Agent
5-38

• Acting as a personal electronic assistant to
  collect, send, and prioritize electronic
  information such as e-mail.
• Finding and retrieving information from a
  database.
• Finding and retrieving information across
  networks.

42
How Does an Intelligent Agent Do That?
Intelligent Agent
5-39
An intelligent agent (usually) combines the
capabilities of two or more of the most modern
software technologies such as expert systems,
neural networks, genetic algorithms,
object-oriented programming, and so on.
43
AI
5-40
Starting Information Expert? know-how Acceptable
patterns Set of possible solutions Your
preferences
AI System Expert Systems Neural
Networks Genetic Algorithms Intelligent Agents
Problem Type Diagnostic or prescriptive Identifi
cation, classification, prediction Optimal
solution Specific and repetitive tasks
Based On Strategies of experts The human
brain Biological evolution One or more AI
techniques
44
AI SYSTEMS CAN BE COMBINED WITH DECISION SUPPORT
SYSTEMS
IT Brainpower
5-41

• DSSs can incorporate one or more expert systems
  creating an intelligent DSS, a DSS/ES, or a
  knowledge-based DSS.
• Neural networks can be combined with GISs to
  apply pattern recognition capabilities to spatial
  information.

45
A HYBRID INTELLIGENT SYSTEM
IT Brainpower
5-42
is an IT system which combines two or more
artificial intelligence systems.

• Neural networks and expert systems can be
  combined to create a ?rainable?expert system.
• Neural networks and genetic algorithms can be
  combined to train neural networks.

46
TO SUMMARIZE
5-43

• IT can help you be an effective decision maker by
  assisting you in decision-making tasks with
• Decision support systems (DSS)
• Group decision support systems (GDSS)
• Geographic information systems (GIS)

47
TO SUMMARIZE
5-44

• IT can help you be an effective decision maker by
  performing tasks for you with AI using
• Expert systems
• Neural networks
• Genetic algorithms
• Intelligent agents

48
TO SUMMARIZE
5-45

• To benefit from the decision support capabilities
  of IT you must know
• The nature of the problem
• What type of decision support tools can help
• How to apply the decision support tools to the
  problem.