Chapter - 7
Complex Decisions and Artificial
Complex Decisions & Artificial Intelligence
Decision of data and model.
Symbolic & Expert decisions
Numeric Knowledge made by
If income > 20,000
or expenses < 3000
and good credit history
or . . .
Then 10% chance of default
Decision Support System and
goal help user make decision provide expert advice
method data - model - presentation asks questions,
applies rules, explains
type of general, limited by user narrow domain
Expert Ssystem Example: Bank loan
the Loan Evaluation System.
What is the purpose of the loan? car
How much money will be loaned? 10,000
For how many years? 5
The current interest rate is 10%.
The payment will be $212.47 per month.
What is the annual income? 24,000
What is the total monthly payments of other loans? Why?
Because the payment is more than 10% of the monthly income.
What is the total monthly payments of other loans? 50.00
Comment from DSS
The loan should be approved, there is only a 2% chance of default.
Decision Tree (bank loan)
monthly income? No
Yes total < 30%
Credit monthly income?
the loan the loan
• United Airlines GADS: Gate Assignment
• American Express Authorizer's Assistant
• Stanford Mycin: Medicine
• DEC Order Analysis + more
• Oil exploration Geological survey analysis
• Auto/Machine repair (GM:Charley) Diagnostic
ES Problem Suitability
• Narrow, well-defined domain
• Solutions require an expert
• Complex logical processing
• Handle missing, ill-structured data
• Need a cooperative expert
• Repeatable decision
• ES Shells Rules
– Jess and
– Exsys trees backward
• Custom Programming by designer by ES shell
– LISP Maintained by expert system shell
Expert Knowledge seen by user
(for (k 0 (+ 1 k) )
exit when ( ?> k cluster-size) do
(for (j 0 (+ 1 j ))
exit when (= j k) do
engineer (connect unit cluster k output o -A
to unit cluster j input i - A ))
Programmer ... )
Custom program in LISP
Some Expert System Shells
– Originally developed at NASA
– Written in C
– Available free or at low cost
– Written in Java
– Good for Web applications
– Available free or at low cost
– Commercial system with many features
Limitations of ES
• Fragile systems • Conflicting experts
– Small environmental
changes can force revision – With multiple opinions,
of all of the rules. who is right?
• Mistakes – Can diverse methods
– Who is responsible? be combined?
• Unforeseen events
• Multiple experts?
• Knowledge engineer? – Events outside of
• Company that uses it? domain can lead to
• Vague rules nonsense decisions.
– Rules can be hard to – Human experts adapt.
define. – Will human beginner
recognize a nonsense
• A collection of a documents and data
– Created by experts
– With links to related topics
– Highly organized groupware
• Emphasizing context
• Example—business decisions
– Store problem, all notes, decision factors, comments
– Future problems, managers can search the database
and find similar problems
– Better and more efficient decisions if you know the
original problems, discussions, and contingency plans
• Main problem—convincing everyone to enter and
update the documents
ArtificiaI Intelligence - Research Areas
• Computer Science • Natural Language
– Parallel Processing – Speech Recognition
– Symbolic Processing – Language Translation
– Neural Networks – Language
• Robotics Applications Comprehension
– Visual Perception • Cognitive Science
– Tactility – Expert Systems
– Dexterity – Learning Systems
– Locomotion & – Knowledge-Based
Neural Network: Pattern recognition
Some of the connections
Incomplete Sensory Input Cells
Machine Vision Example
The Department of Defense has funded Carnegie Mellon
University to develop software that is used to automatically drive
vehicles. One system (Ranger) is used in an army ambulance
that can drive itself over rough terrain for up to 16 km. ALVINN is
a separate road-following system that has driven vehicles at
speeds over 110 kph for as far as 140 km.
• Look at the user’s voice command:
• Copy the red, file the blue, delete the yellow
• Now, change the commas slightly.
• Copy the red file, the blue delete, the yellow
I saw the Grand Canyon flying to New York.
Subjective (fuzzy) Definitions
cold hot temperature
Moving farther from the reference point
increases the chance that the temperature is
considered to be different (cold or hot).
DSS, ES, and AI: Bank Example
Decision Support System Expert System Artificial Intelligence
Loan Officer ES Rules Determine Rules
Income What is the monthly income? Data/Training Cases
Data Existing loans 3,000 loan 1 data: paid
Credit report What are the total monthly loan 2 data: 5 late
payments on other loans? 450 loan 3 data: lost
loan 4 data: 1 late
Lend in all but worst cases How long have they had the
Monitor for late and missing current job? 5 years
Name Loan #Late Amount
Output Brown 25,000 5 1,250
Neural Network Weights
Jones 62,000 1 135 Should grant the loan since there
Smith 83,000 3 2,435 is only a 5% chance of default.
Evaluate new data,
• Independent book trip.
• Networks/Communication Software agent
– Negotiate Resorts
• What is intelligence?
– Ability to handle unexpected events?
• Can machines ever think like humans?
• How do humans think?
• Do we really want them to think like us?
Appendix: E-Mail Rules -
Folders make it
easy to organize
and handle your
Simple rules from
the Tools +
directly to the
The Tools + Rules
Wizard makes it easy to
create rules. Begin with
a blank rule.
Set the Conditions
Set the Actions
A sample rule to handle
unsolicited credit card
Choose an action.
You can choose multiple
actions, but be careful.
The marking options are
Rules can have
exceptions. For example,
you might want to delete
unless one has your name
Rule Sequences: Decision Tree
Message from Expenses Folder
Rule 1 Accounting Set expenses category
Rule 2 From boss,
Subject: Expenses Action: Mark important
best firm in the
• Do a situation analysis, self-evaluation and competitor
analysis: both internal and external; both micro and
• Concurrent with this assessment, objectives are set.
– vision statements (long term view of a possible future),
– mission statements (the role that the organization gives itself in
– overall corporate objectives (both financial and strategic),
– strategic business unit objectives (both financial and strategic),
and tactical objectives.
• These objectives should, in the light of the situation
analysis, suggest a strategic plan. The plan provides the
details of how to achieve these objectives.
• Allocation of resources (financial, personnel,
time, technology support)
• Establishing a chain of command or some
• Assigning responsibility of specific
tasks/processes to individuals or groups
• Management of the process. Monitoring results,
comparing to benchmarks, evaluating the
• Acquiring the necessary resources, developing
the process, training, process testing,
documentation, and integration with inheritance
Michael Porter’s Five Forces Model
Bargaining Power Rivalry Among
of Suppliers Bargaining Power
Existing Competitors of Buyers
Threat of Substitute
Products or Services
The Porter 5 forces analysis is a framework for industry analysis and business
strategy development developed by Michael Porter in 1979. It uses concepts
developed in Industrial Organization (IO) economics to derive 5 forces that determine
the competitive intensity and therefore attractiveness of a market.
Driving Force : Competition
• Competition is increasing in many industries.
• Competition encourages firms to hold down costs,
provide more variety, and provide new and better service
•According to microeconomic theory, no system of
resource allocation is more efficient than pure
•The greater selection typically causes lower prices for
the products compared to what the price would be if
there was no competition (monopoly) or little competition
parts parts parts
Chain supplier supplier supplier
supplier supplier supplier
distributor distributor distributor
retail store retail store retail store retail store
Methods to Gain
Competitive To Entry
Improved Control Of
Consumer Consumer Consumer
Barriers to Entry
• Economies of Scale (size)
• Economies of Scope (breadth)
• Product Differentiation
• Capital requirements
• Cost Disadvantages (independent of size)
• Distribution Channel Access
• Government Policy
Competitive Advantage with IS
• Barriers to Entry • Lower Production Costs
– Additional costs of – IS to cut costs.
creating an information • Product Differentiation
system. – Add new features or create
• Distribution Channels new products with IT.
– Prevent others from • Quality Management
entering the industry. – Monitoring production lines
and analyzing data.
• Switching Costs • Value Chain
– Consumers invite – Expanding forward or back
learning and data the value chain to find
transfer costs. greater profits.
Value-adding activities of an organization.
The "primary activities" include: inbound logistics, operations (production),
outbound logistics, marketing and sales, and services (maintenance).
The "support activities" include: administrative infrastructure management, human
resource management, R&D, and procurement.
The costs and value drivers are identified for each value activity. Its ultimate goal is
to maximize value creation while minimizing costs.
Process Innovation Suppliers Production
Service Sales and
Business Strategies leadership
Corporate Strategy Development and Priorities - Differentiation
monitor • expectations • strengths - Innovation
rivals • goals • weaknesses - Linkages
• rivalry • opportunities
• critical success factors - Re-engineering
Performance Measures Decentralization
- ROA - ROI
Market Measures - EPS - Growth
- Market share - Subjective
- Growth & Implementation
Business Operations & Rules
Existing Data and IS
Search for Innovation with IS
• Research • Manufacturing
– Analysis & modeling, – Mass customization, links
to customers & suppliers,
project management, quality monitoring, expert
work group support, systems for maintenance,
databases, decision production databases,
support. business integration.
• Engineering & Design • Logistics & Supply
– Just-in-time linkages,
– CAD/CAM, testing, forecasts, models, links for
networks, work group design, transaction
Search for Innovation with IS
• Marketing • Service
– Frequent buyer database, – Phone support, GIS
target market & media locators, scheduling, ES
analysis, survey design diagnostics, databases.
and analysis, multimedia • Management
promotion design, links to
customers and designers. – EIS, e-mail, bulletin boards,
decision support systems,
• Sales & Orders personal productivity tools,
– Portable computers for work group support
sales, ES for order – Links to service providers
customization, work group • Accountants
tools for customer support. • Consultants
• Lawyers, . . .
Research with IS
• Analysis and models
• Statistical analysis of data
• Project management and
• Work-group collaboration and
Engineering and Design with IS
• Integrated design database
• Production databases and model testing
• Expert Systems for manufacturability
• Work group communication
Manufacturing with IS
• Links to customers
• Links to suppliers
• Mass customization
• Diagnostic Expert Systems
• Quality monitoring and control
Logistics and Supply with IS
• Just-In-Time Inventory and EDI
• Configuration and design
• Searching for availability, pricing, . . .
Marketing with IS
• Frequent buyer databases
• Point-of-Sale and trends
• Statistical analysis of data
• Geographic Information Systems
• Links to external marketing agencies
• Multimedia development of promotions
Sales and Orders with IS
• Sales force automation, hand-held
• Customer Internet access
• Expert Systems for product and option
• Expert Systems for configuration and
• Front-line support: ES, e-mail, work
Service with IS
• Portable computers for service anywhere
• Databases (e.g., customer service)
• Location monitoring of service personnel
• Product internal, automatic diagnostics
• Expert System diagnostic tools
Management with IS
• Executive Information Systems
• Simulation (and rivalry games)
• Links to external partners (accounting, law, . . .)
• Electronic conferencing
• Work group communication, e-mail
• Standardization, Modularization, Franchises
• Knowledge Workers
• Client-server instead of hierarchical computing
• Product Differentiation • Cost Leadership
– Skills & Resources – Skills & Resources
• Strong marketing. • Continued capital
• Product engineering. investment.
• Basic research. • Process engineering.
• Distribution channel • Continuous quality
– Organization Requirements • Tight supervision of costs.
• Internal coordination. • Products designed for low
• Incentives for innovation.
• Low cost distribution.
• Resources to attract skills.
– Organization Requirements
• Tight cost controls.
• Competitors imitate.
• Frequent control reports.
• Customers do not accept.
• Highly structured org.
• Cost is too high.
• Incentives based on
• Cost Leadership • Customer-Supplier Links
– Risks – Skills & Resources
• Influence with partners
• Competitors imitate.
• Communication channels
• Technology changes.
• Standards or agreements.
• Lose production or – Organization Requirements
• Flexibility to respond to
• Service culture.
• Ability to adapt to
• Security threats.
• Changing standards.
• Competitors copy with
Transaction Network & link Experimental
Processing sales people technology
& global links
Network & DSS Link to suppliers
Changing Industry &
New services Customer
Industry 1 Customer Industry 2
(expands into (new
industry 2) competitor)
Customer Same technology
Security Need to control access.
Need to worry about network
interceptions and hackers.
Data we wish to share.
Data we want to protect. Customer
• Hundreds of dot-com firms failed in 2001 and 2002
• Most relied on pure Internet revenue. Outsourcing
production and shipping.
• Most relied on advertising revenue—often revenue from
other dot-com firms.
• Many believed in the importance of being first to market
and becoming the biggest, best-known firm in a niche
• Many believed that it was not necessary to make a profit
on sales. Money from advertising and stock sales would
be sufficient to keep the firm alive until the world
• Most were wrong.
• Does the project fit with business goals and management style?
• Does the project improve the competitive position of the firm?
• How long will any competitive advantage last?
• What value or reward is created by the system?
• What level of technology is needed to create the system?
– leading edge
• What is the probability of technical success?
• What is the probability of commercial success?
• What are the costs involved in creating the system?
– Additional capital, marketing and management
Small business/ Large business and
The Internet CRM
Web hosting and information
Forms of Electronic Commerce
Business B2B B2C
Consumer C2B C2C
Minimal Auction sites (eBay)
examples, But many of these
possibly reverse are dominated by
auctions like small business
Production Chain Disintermediation
• Pre-Purchase • Purchase
– Transmission security.
– Static data sites.
– User identification.
– Product selection.
• Product specifications.
– Payment validation.
– Order confirmation.
• Problem tracking.
– Interactive sites. • Sales leads.
• Configuration. – Resolve problems.
• Compatibility. – Answer questions.
• Complex pricing. – Product evaluation.
• Tracking customers.
Simple Static HTML Website
Main Web Page
Category 1 Category 2 Category 3
Product photo Product photo Product photo
… … … … … …
Product 1 Product 2 Product 3 Product n
Description Description Description Description
Price Price Price Price
Photo Photo Photo Photo
Simple Website with Buy Me
Merchant Web site Button
Card Processor Site
Product Buy Me
Description Shopping Cart Credit Card Data
Price Item Price Name
… … Address
Check Out Card Number
Credit Card Processing
Online Pay registrar
Card (Verisign) $35/year for
Processor domain name.
CC Confirm Pay CA (Verisign)
Issuing Data $250/year for
Bank Merchant certificate
Web Server Pay card processor
Digital Merchant bank fixed
Certificate fee and value fee:
• Uncertain price
• Can set reserve price
• Good for unique items
• Efficiency depends on
– Full information
– Adequate number of
Cameras, Digital, Brand
Description Vendor 3
Scanned image Product search
Contact info Transaction Processing Choose vendor
Amazon.com handles credit Pay for item
Sends order info to merchant
Merchant ships item to consumer
Web Commerce Servers
Your Web site
Commerce Server Shell
Web/Commerce Hosting Company
Application Service Provider
Businesses that lease the use of the application