제2주 강의 - 부산대학교 인공지능

Download Report

Transcript 제2주 강의 - 부산대학교 인공지능

제2주 강의
지능적 프로그램의 특징
인공지능 연구에 대한 접근 1
• A branch of computer science to
automate intelligent behavior
• Study on computational techniques to
perform tasks requiring intelligence
• Generating representations and
procedures that automatically and
autonomously
인공지능 연구에 대한 접근 2
• Study on explaining and emulating
intelligent behavior in terms of
computational processes
• Designing intelligent computer
systems that can understand
languages, learn, reason and solve
problems
Problems to be solved
•
•
•
•
•
GPS (general problem slover)
Commonsense reasoning
Vision and image processing
Expert system
Language understanding and
speaking
• Learning
Questions about AI
• What are the underlying assumptions
about intelligence?
• What kinds of techniques will be
useful for solving AI problems?
• At what level if at all can human
intelligence be modelled
• When will it be realized ?
Underlying Assumption
• Symbolic approach
*** 이전 시간에 일부 설명
Knowledge
• Capture generalizations rather than
separate representation
• Be understood by people providing it
• Be easily modified to correct errors and
changes in real conditions
• Be widely used even if it is incomplete and
inaccurate
예
• Tic-Tac-Toe
 단순한 접근 ::: 19683 elements
 규칙화 … blank: 2, X:3, O:5; 순서
를 1,2 3, …; Win 검증 3*3*2 (X win),
5*5*2 (O win)
 board position을 이용하면서 규칙
화, 가능성을 검증하여 움직임을 결정
AI Techniques
• Search
• Problem characteristics
• 예
missionary carnival problems,
simple question-answering system
water jug problem
constraint satisfaction
Problem Characteristics
• Can solution steps be ignored or at least undone
if they prove unwise?
• Is the problem’s universe predictable?
• Is a good solution obvious?
• A large amount of knowledge is required or
important to constraint the search?
• Is the problem decomposable to independent
easier sub-problems?
• Does the solution of the problem require
interaction?
Question Answering
• TEXT
Many went shopping for a new coat.
She found a red one she really liked.
When she got home, she discovered
that it went perfectly with her favorite
dress.
Questions
• What did Mary go shopping for?
• What did Mary find that she liked?
• Did Mary buy anything?
• Eliza
: Template matching
: Slot filling
Agents
• Acting humanly ?  Turing test
• Thinking Humanly : The cognitive modeling
approach
• Thinking rationally : The laws of thought
approach :: correct inference
• Acting rationally :: The rational agent approach
 not thinking rationally but acting rationally :
stove에서 손을 뗄 때
 evolution, neural network
숙제
• 교재 1.2, 1.3, 1.4, 1.7, 1.8, 1.9, 1.10
• Turing test가 지능을 측정하는 방법이 될 수
있는가? 있다면 그 입장에서 논리를 전개하고,
그렇지 않다면 그 입장에서 논리를 전개하라.
• 음식점에 들어가서 나오는 과정을 행위의 단
위를 나누어서 기술하라. 자신이 기술한 과정
에서 사용된 기본 연산(operation)을 기술하
라. 연산이 다른 음식점에서도 사용될 수 있는
지 보여라. 그 연산을 일반화하라. 마지막으로
음식점에서 일어나는 과정을 일반화해보라.