Artificial Intelligence: Introduction - Indian Institute of Technology ...

23 downloads 255899 Views 343KB Size Report
CSE, IIT Kharagpur. What is AI? ▫ Turing Test (1950). ◇ The computer is interrogated by a human via a teletype. ◇ It passes if the human cannot tell if there is a ...
Artificial Intelligence

Course: CS40002 Instructor: Dr. Pallab Dasgupta

Department of Computer Science & Engineering Indian Institute of Technology Kharagpur

What is AI? „

Turing Test (1950)  

The computer is interrogated by a human via a teletype It passes if the human cannot tell if there is a computer or human at the other end

• Sufficiency: The Chinese Room Argument CSE, IIT Kharagpur

2

The ability to solve problems „

Search: Efficient trial-and-error ‹ ‹ ‹

Enormous computational complexity Space-time trade-offs Use of domain knowledge – heuristics

Linear Prog.

Integer Prog.

Dynamic Prog.

Heuristic Search

Evolutionary Algorithms

During 1985-1995 computation became free CSE, IIT Kharagpur

3

Knowledge and Deduction „ „ „ „

How to store and retrieve knowledge? How to interpret facts and rules, and be able to deduce? The gap between knowledge and realization Logics of knowledge

Knowledge Based Systems

Expert Systems

Automated Theorem Provers

Formal Verification

• The knowledge base may be huge • Between 1990 – 2000 storage became free CSE, IIT Kharagpur

4

The ability to learn „

Can we learn to solve a problem better? ‹ ‹ ‹

Learning the answers Learning the rules of the game Learning to plan

„

Belief networks

„

Perceptrons and Neural networks

CSE, IIT Kharagpur

5

What then is AI? Human Computer interaction

Automated Problem Solving

Machine Learning Logic and Deduction

Computer vision

NLP

Robotics

In this decade, communication will become free CSE, IIT Kharagpur

6

Fundamentals „

The notion of expressing computation as an algorithm

„

Godel’s Incompleteness Theorem (1931): ‹

In any language expressive enough to describe the properties of natural numbers, there are true statements that are undecidable: that is, their truth cannot be established by any algorithm.

CSE, IIT Kharagpur

7

Fundamentals „

Church-Turing Thesis (1936): ‹

‹

„

The Turing machine is capable of computing any computable function This is the accepted definition of computability

The notion of intractability ‹ ‹

NP-completeness Reduction

CSE, IIT Kharagpur

8

Course Outline „

Problem solving by search ‹ ‹ ‹

„

Logic and deduction ‹

„ „ „ „

State space search, Problem reduction search, Game playing First-order logic, Temporal logic, Deduction

Planning Reasoning under Uncertainty Learning Additional Topics

CSE, IIT Kharagpur

9

References „

Artificial Intelligence – A Modern Approach -- Stuart Russell and Peter Norvig

„

Principles of Artificial Intelligence -- N J Nilsson

„

Heuristics -- Judea Pearl

CSE, IIT Kharagpur

10