Artificial Intelligence : An Overview - PowerPoint PPT Presentation

1 / 37
About This Presentation
Title:

Artificial Intelligence : An Overview

Description:

Title: (Intelligence Technology) Author: Last modified by: Created Date: 6/17/1995 11:31:02 PM – PowerPoint PPT presentation

Number of Views:331
Avg rating:3.0/5.0
Slides: 38
Provided by: 6649912
Category:

less

Transcript and Presenter's Notes

Title: Artificial Intelligence : An Overview


1
Artificial Intelligence An Overview
  • ? ? ?

2
????(??? ??)
  • ??? ????? ??? ??? ?? ?? ???? ???? ?? ??/??
  • ??? ??? ?? ? ?
  • ???? ? ? ???? ??? ??/??
  • ???? ??? ??? ??/??
  • Can machine think ?

3
????(????? ??)
  • ????? ??? ??? ???? ?
  • ????? ??? ??
  • ?? ??(Artificial Mind)

4
4 Views of Definition
Rationality
Human-likeness
Think like Human
Think Rationally
Thinking Acting
Cognitive Modeling Approach
Laws of Thought Approach
Act like Human
Act Rationally
Turing Test Approach
Rational Agent Approach
5
??? ???
? ? ? ? ? ? ? ? ?
??, ?? ??, ??
? ??
??
6
???????? ?
  • ???????
  • ???????
  • Intelligent Building System
  • ?? ????, Softbot
  • ??? ????
  • 3D ??? ??
  • ????, ???? ???

7
Turing Test
  • ?? ??? ????? ??
  • Imitation Game
  • Intelligent as much as Human
  • Is dog intelligent ?
  • Any man-made system passed Turing Test ?

8
Imitation Game
9
?? ??? ???
  • ??? ???? 50 ??? ??? ???? ???? ??? ??
  • ???
  • ??? ??
  • ??? ??
  • ?? ??
  • ???? ???
  • ??????
  • ?? ??
  • ?????
  • ??? ????
  • ??? ??
  • ????
  • .....

10
History ??? (1943-1956)
  • McCulloch Pitts (1943)
  • Propose Artificial Neuron Model
  • Reacting to the stimulus
  • Learning
  • Shannon (1950) Turing (1953)
  • Wrote Chess Program
  • Minsky Edmonds (1951)
  • Built 1st NN computer
  • McCarthy (1956)
  • Named the field by Artificial Intelligence

11
History ??? (1952-1969)
  • Early enthusiasm, Great expectations
  • Newell Simon (CMU)
  • GPS
  • 1st program to embody thinking humanly
  • Rochester et. al. (IBM)
  • Geometry Theorem Prover
  • 1st AI Program
  • Lisp by McCarthy (MIT, 1958)
  • Microworlds Approach
  • Perceptrons by Rosenblatt (1962)

12
History ??? (1966-1974)
  • Lack of enough knowledge
  • the spirit is willing but the flesh is weak is
    translated to the vodka is good but the meat is
    rotten
  • Intractability of problem
  • NP-complete problem
  • Illusion of unlimited computational power
  • Fundamental limitation on the basic structures
  • perceptrons could not handle XOR

13
History ???? ???? ?? (1969-1979)
  • Focusing on the domain knowledge
  • DENDRAL (1969, Buchanan)
  • inferring molecular structure from mass
    spectrometer information
  • 1st successful knowledge-intensive system
  • Heuristic Programming Project (Stanford)
  • new methodology of Expert System
  • Knowledge Representation Schemes
  • Prolog in Europe, PLANNER in US
  • Frames (1975, Minsky)

14
History ??? (1980-1988)
  • R1 (DEC, McDormatt, 1982)
  • 1st commercial expert system, 10million/yr
  • By 1988, DEC deployed 40 expert systems
  • Nearly almost major US corporation
  • 5th Generation Project of Japan (1981)
  • Industry
  • S/W Tools for expert system
  • H/W companies building optimizing for Lisp
  • industrial robotics vision systems
  • few million , 1980 -gt 2 billion, 1988

15
History NN?? ?? (1986-??)
  • Reinvent back-propagation learning algorithm
  • applied to many learning problems
  • Parallel Distributed Processing collection
    (Rummelhart McClelland, 1986)
  • Widely used in learning
  • character recognition
  • data mining

16
History ?? ?? (1987 - ??)
  • HMM (Hidden Markov Models)
  • speech technology
  • handwritten character recognition
  • Formalization of Planning
  • factory scheduling
  • Probablistic Reasoning
  • accept probality and decision theory in AI
  • Back to Strong AI SOAR(Newell, 1990)

17
State of the Art
  • 1st program grandmaster HITECH (Berliner, 1989)
  • PEGASUS NLP traveling agent (Zue, 1994)
  • MARVEL real-time expert system for spacecraft
    (Schwuttke, 1992)
  • 55mph automated driving (Pomerleau, 1993)

18
????? ????
  • ??? ??
  • ??? ? ??? ?
  • ?? ??(Performance)
  • ????? ??
  • ??? ??? ???? ?? ?
  • Simulation of Behavior

19
????? ?? ???
  • u
  • Knowledge-based Approach
  • u
  • Data Driven Approach
  • u

20
????? ???
  • u
  • Represent Human knowledge as symbol combination
  • u
  • Knowledge Acquisition and Representation
  • u
  • Logic, Expert System, Fuzzy Logic
  • u

21
Data Driven Approach
  • u
  • Extract common characteristics from collected
    examples
  • u
  • Training
  • u
  • Statistical Methods, Artificial Neural Network

22
Generality vs Power
  • ???? ???? ?? ???? ??
  • General Problem Solver
  • ??? ?? ??
  • Complexity Toy Problems Only
  • Power? ??? Specialized Approach?
  • Knowledge Based Approach
  • ???? Expert Systems

23
?????? (????)
  • Symbolic Programming
  • ????
  • Search Planning
  • Automated Reasoning
  • Machine Learning
  • Artificial Neural Net
  • Genetic Algorithm
  • ...

24
?????? (????)
  • ???? ??
  • ?? ? ?? ??
  • ????
  • Uncertainty Modeling
  • ??? ???
  • Virtual Reality
  • ..

25
Symbolic Programming
  • Program as Representation of world
  • Symbol as basic element of representation
  • atom, property, relationship
  • Symbolic Expression as method of combination
  • LISP for Symbolic programming
  • Object-Oriented Concept

26
????
  • ??? ??? ??? ?
  • ??? ??? ??? ?
  • declarative vs procedural
  • ?? ??? ?
  • explicit vs (implicit inference)
  • logic, frame, semantic net, script
  • ??? ??? ??

27
Search Theory
  • ??? ??
  • ??? ? ?? ??? ??/??
  • ??????? ???? ?????
  • ????? ???(Heuristic Search)
  • ??? ?? ??(Hill Climbing Method)
  • ??? vs ??? ?? ?
  • Genetic Algorithm, Simulated Annealing

28
Automated Reasoning
  • Qualitative Reasoning
  • ??? ??? ??
  • Non-monotonic Reasoning
  • ??? ?? ? ??? ?
  • Plausible Reasoning
  • ??????? ???? ??? ??
  • Case-based Reasoning
  • ??? ??

29
Machine Learning
  • ??? ??? ?? ??? ??? ??
  • ??? ??? ??? ?
  • ??? ???? ??? ??? ??? ?
  • ??? ??? ??? ?? ??
  • ??? ??
  • ??? ??? ??
  • Parameter Adjustment
  • Data Mining?? ??? ??

30
?????
  • ????? ????
  • ?? ??? ??, ??? ??
  • ??? ?? ?? ?? ??

X1
w1
w2
X2
F(X1, X2, , Xn)
S
. . .
wn
Xn
31
????? (Neural Network)
  • ?? ???? ??
  • Error-back-propagation ????????
  • ??? Functional Mapping? ?? ???
  • Sensory Data Processing? ??
  • Old Horse on the race again
  • ??? ???, graceful degradation
  • Symbolic Grounding

32
Neural Network Classifier
Input layer
Hidden layer
Output layer
33
Genetic Algorithm
  • ??? ??? ?? ???
  • ??? ??? ???
  • ?? ??? ??,
  • ??? ?? (??, ????)
  • ?? ??
  • ???? ??? ??
  • ????? ??? ??

34
????? ??? ?? Data Mining
????? ?? ?? ??
???
??
????
  • ????
  • Point of Sale
  • ATM
  • ????
  • ????
  • ??
  • ????
  • ????
  • ??????
  • A?? ???? 80? B??? ????
  • ????? ??? ???? 6??? ??
  • A??? ?? ??? B??? 2?
  • ?? ??? ??? ??
  • ????? ?
  • ??? ??
  • ??? ?? ??? ?
  • ????? ????? ?
  • ??? ?? ???? ?
  • ??? ?

35
????? ??
  • ???? ??? ????
  • OCR, ICR, Symbolic Algebra, Machine Translation,
    Many Expert systems, Planning systems
  • ??? ??, ????? ??? ??? ??
  • Programming Language, DataBase, Operating System

36
????? ??
  • ? ??? ??? ?? ?????
  • ???? ??? ???
  • Web auto translation system
  • ??, ?? ?? Interface Package
  • ?? Paradigm ??
  • Symbolic Processing Neural Processing

37
????? ??
  • ????? ?? ?
  • AI in everywhere, AI in nowhere
  • ?? ??? ?????
  • Ubiquitos Computing
Write a Comment
User Comments (0)
About PowerShow.com