Transcript Decision support system

CHAPTER 4
DECISION SUPPORT AND
ARTIFICIAL INTELLIGENCE
Brainpower for Your Business
AN NFL TEAM NEEDS MORE THAN
ATHLETIC ABILITY




The Patriots football team is a very
successful one
The team uses a decision support system
to analyze the opposition’s game
The software breaks down the game day
video into plays and player actions
With this information the Patriots can
better formulate their strategy
INTRODUCTION

Computer-aided decision support
DECISIONS, DECISIONS,
DECISIONS

Phases of decision making




Intelligence
Design
Choice
Implementation
Four Phases of Decision Making
Types of Decisions

Structured decision

Semi-Structured decision

Nonstructured decision
What Job Do I Take?
Types of Decisions You Face

Recurring decision

Nonrecurring (ad hoc) decision
DECISION SUPPORT SYSTEMS


Decision support system (DSS)
Help you analyze, but you must know how
to solve the problem, and how to use the
results of the analysis
Alliance between You and a DSS
Components of a DSS

Model management component

Data management component

User interface management
component
Components of a DSS
GEOGRAPHIC INFORMATION
SYSTEMS



Geographic information system (GIS)
Spatial information is any information in
map form
Used to analyze information, generate
business intelligence, and make decisions
Zillow GIS Software for Denver
ARTIFICIAL INTELLIGENCE

Artificial intelligence (AI)

Types of AI systems used in business
1.
2.
3.
4.

Expert systems
Neural networks
Genetic algorithms
Intelligent agents
AI systems deliver the conclusion (rather
than helping you analyze the options)
EXPERT SYSTEMS

Expert (knowledge-based) system

Used for


Diagnostic problems (what’s wrong?)
Prescriptive problems (what to do?)
Traffic Light Expert System
Components of an Expert
System
1.
Information Types



2.
Domain expertise
“Why?” information
Problem facts
People



Domain expert
Knowledge engineer
Knowledge worker
Components of an Expert
System
3.
IT Components
• Knowledge acquisition
•
Explanation module
•
User interface
•
Inference engine
•
Knowledge base
Developing and Using an Expert System
What Expert Systems Can and
Can’t Do
 An
expert system can
 Reduce
errors
 Improve customer service
 Reduce cost
 An
expert system can’t
 Use
common sense
 Automate all processes
NEURAL NETWORKS

Neural network (artificial neural
network or ANN)
The Layers of a Neural Network
Neural Networks Can…
 Learn
and adjust to new circumstances on
their own
 Take part in massive parallel processing
 Function without complete information
 Cope with huge volumes of information
 Analyze nonlinear relationships
Fuzzy Logic

Fuzzy logic – a mathematical method of
handling imprecise or subjective information
 Used to make ambiguous information such as
“short” usable in computer systems
 Applications
Google’s search engine
 Washing machines
 Antilock breaks

GENETIC ALGORITHMS



Genetic algorithm
Takes thousands or even millions of
possible solutions, combining and
recombining them until it finds the optimal
solution
Work in environments where no model of
how to find the right solution exists
INTELLIGENT AGENTS

Intelligent agent




Information agents
Monitoring-and-surveillance or predictive agents
Data-mining agents
User or personal agents
In Class Team Activity
A music owner wants to keep enough hot CDs in stock, so
as not to lose sales and disappoint customers. CDs that
are not sold for some time, go on the sale table, and are
sold at below cost, if they sell at all. She wants to
design a DSS to try and predict how many copies of
each future CD she should order, and for how long she
should keep them on a rack, before putting them on
sale. As Strategic Consultants to her, our team needs to
list some of the considerations that would go into such a
system. What kind of DSS should she use?
Hint: Some examples: how often does her customers base
change, previous sales of particular types of music with
Copyright 2004
her customer base
The McGraw-Hill
4-28
Management Information Systems
for the Information Age
Companies, Inc.
All rights reserved