Artificial Intelligence

1
Artificial Intelligence
2
Artificial Intelligence

• Systems deal with intellectual and logical
  operations
• utilizes programming algorithms to simulate
  thought processes
• reasoning that produce behavior similar to humans

A successful implementation of AI could be
tested by using a Turing Test
3
Turing Test

• Alan Turing in 1950s
• No physical contact between human the machine
• If the machine can "fool" human, it is
  intelligent.

4
More about AI

• made up of multiple intelligent agents that have
  specific functions within its own domain.
• The agent is typically programmed to perceive
  certain changes
• have sensors that are used to perceive changes
  utilizes effectors to produce actions
• can learn from those perceptions by testing its
  actions on the environment

5
Application of AI

• Computer Vision
• Natural Language Processing
• Pattern Recognition
• Robotics
• Expert System

6
Computer Vision

• Emulation of biological vision using computers
  machines
• Application
• Robot vision
• Object recognition
• Animation automation

7
Natural Language Processing

• Aim of research
• designing computers that can understand our
  natural languages.
• Limited-scope of natural language processors
• The commands are restricted to some specific
  patterns. The users have to speak slowly in a
  word by word manner.

8
Pattern Recognition

• The mechanism of translating written words into
  computer-readable forms.
• Pattern recognition is usually done in three
  steps
• receive the text, analyze the characters, and
  compare the patterns.

9
1st step Receive The Text

• done by detecting the presence of the stylus that
  emits a faint signal from its tip. As the user
  writes, the screen grows darker on where the
  stylus touches it.
• In an optical character recognition (OCR) system,
  it is done by using an optical scanner to scan
  the text.

10
2nd step Analyze the characters

• The system breaks each character into its
  horizontal, vertical, and diagonal elements. No
  two characters have exactly the same number of
  elements of each type.

11
3rd step Compare the patterns

• It is to compare the pattern of those elements
  with the patterns in a set of stored images of
  characters.
• The system narrows the choices until the best
  match is found.

12
Robotics

• It is the technology of building and using robots
  with artificial intelligence and computer
  controlled human-like capabilities.
• Example of Application
• used in factories
• Advantage
• reliable, uncomplaining, never getting tired or
  bored.
• Disadvantage
• Scientists need to design a number of arms for
  doing specific jobs.

13
(No Transcript)
14
Expert System

• Provide quick and inexpensive access to expert
  knowledge.
• Accepts questions from users and can evaluate,
  draw conclusions, and make recommendations based
  on a huge database of information in the
  particular field.
• Example Medical diagnosis expert system

15
Expert System

• Knowledge Base
• contains the rules, data, and relationships on a
  particular field
• Inference Engine
• a special program that can, with reference to the
  knowledge base, evaluate users' questions, draw
  conclusions and make recommendations.
• Explanation Module
• a special program that enables the expert system
  to explain why it arrives at a certain conclusion
  or makes certain recommendations.
• Natural Language Processor
• makes possible the expert system to communicate
  with users in a natural manner

16
Disadvantages in expert systems

• limited by the knowledge base
• cannot really " think" like human
• Human experts may violently disagree on many
  subjects so that forming definite rules for
  expert systems is impossible
• Many experts cannot express their expertise in
  making a judgment as a set of rules.

17
Reference

• http//www.elearning-engineering.com/technology/ai
  /
• http//www.aaai.org/AITopics/html/turing.html

18
E
N
D