Artificial Intelligence

1
Artificial Intelligence
2

• Definition
• Artificial Intelligence is the study of how to
  make computers do things at which, at the moment,
  people are better.

3

• The Turing Test

According to this test, a computer could be
considered to be thinking only when a human
interviewer, conversing with both an unseen human
being and an unseen computer, could not
determine which is which.
4

• More on AI

Artificial
Real Items Airplanes
Birds Silk Flowers
Flowers Artificial Snow
Snow
5

• AI Major Areas

- Expert Systems - Natural Language Processor
- Speech Recognition - Robotics - Computer
Vision - Intelligent Computer-Aided Instruction
- Data Mining - Genetic Algorithms
6

• Artificial vs. Natural (Human) Intelligence

7

• AI Advantages

1. AI is permanent 2. AI offers ease of
duplication 3. AI can be less expensive than
natural intelligenc 4. AI is consistent 5. AI
can be documented
8

• Natural Intelligence Advantages

1. Natural intelligence is creative. 2.
Natural intelligence uses sensory experience
directly, whereas most AI systems must work
with symbolic input. 3. Human reasoning
is able to make use at all times of a very
wide context experience and bring that to bear
on individual problems, where as AI systems
typically gain their power by having a very
narrow domain.
9

• Characteristics of a Human Experts

- Recognize and formulate the problem -
Solve the problem fairly quickly - Explain the
solution - Learn from experience -
Restructure knowledge - Break rules -
Determine relevance - Degrade gracefully
10

• What Do Experts Know?

It is estimated that a world-class expert, such
as a chess grandmaster, has 50,000 to 100,000
chunks of heuristic information about his/her
specialty. On the average, it takes at least 10
years to acquire 50,000 rules.
11

• Expert Systems

12

• Expert Systems Components

1. Knowledge Acquisition 2. Knowledge Base
3. Inference Engine 4. User Interface 5.
Explanation Facility 6. Knowledge Refining
System
13

• Different Categories of Expert Systems
• Category Problem
  Addressed
• Interpretation Inferring situation
  description from observations
• Prediction Inferring likely
  consequences of given situations
• Diagnosis Inferring systems
  malfunctions from observations
• Design Configuring objects under
  constraints
• Planning Developing plans to achieve
  goals
• Monitoring Comparing observations to plan
  vulnerabilities
• Debugging Prescribing remedies for
  malfunctions
• Repair Executing a plan to
  administer a prescribed remedy
• Control Interpreting, predicting,
  repairing, and monitoring
• system behavior

14

• What Tasks Are ES Right For?
• - Payroll, Inventory
• - Simple Tax Returns
• - Database Management
• - Mortgage Computation
• - Regression Analysis
• - Facts are Known
• - Expertise is Cheap

Too Easy - Use Conventional Software
15

• What Tasks Are ES Right For?
• - Diagnosing and Troubleshooting
• - Analyzing Diverse Data
• - Production Scheduling
• - Equipment Layout
• - Advise on Tax Shelter
• - Facts are known but not precisely
• - Expertise is expensive but available

Just Right
16

• What Tasks Are ES Right For?
• - Designing New Tools
• - Stock Market Forecast
• - Discovering New Principles
• - Common Sense Problems
• - Requires Innovation or Discovery
• - Expertise is not available

Too Hard - Requires Human Intelligence
17

• Problems and Limitations
• of Expert Systems

- Knowledge is not always readily available.
- Expertise is hard to extract from humans. -
ES work well only in a narrow domain. - The
approach of each expert to problem under
consideration may be different, yet correct.
18

• Necessary Requirements for
• ES Development

- The task does not require common sense. -
The task requires only cognitive, not physical,
skills. - There is an expert who is willing to
cooperate. - The experts involved can
articulate their methods of problem
solving. - The task is not too difficult. -
The task is well understood, and is defined
clearly. - The task definition is fairly
stable. - Problem must be well bounded and
narrow.
19

• Justification for
• ES Development

- The solution to the problem has a high
payoff. - The ES can capture scarce human
expertise so it will not be lost. - The
expertise is needed in many locations. - The
expertise is needed in hostile or hazardous
environment. - The system can be used for
training. - The ES is more dependable and
consistent than human expert.
20

• Feasibility Study

A. Financial Feasibility Cost of system
development
Cost of maintenance
Payback period
Cash flow analysis
B. Technical Feasibility Interface
requirements
Network issues
Availability of data and
knowledge Security of confidential
knowledge
Knowledge representation scheme
Hardware/software
availability
Hardware/software compatibility
21

• More on Feasibility Study

C. Operational Feasibility Availability of
human resources
Priority compare to other
projects
Implementation issues
Management and user
support
Availability of experts
Availability of
knowledge
engineers