Artificial Intelligence

Download Report

Transcript Artificial Intelligence

Artificial Intelligence
Presented By:
Abdur Rahim
03-CIVIL-961
Ovais Ali
03-CIVIL-973
Haseeb Khan
03-CIVIL-986
Tajammal Abbas 03-CIVIL-974
Overview

Intelligence
 What is AI
 Branches
 Neural Networks
 Historical overview
 Applications
 Limitations
 Turing test
 Machine learning
 Chess & AI
 AI based movies
 Future of AI
Intelligence
 General
mental capability to reason
 Solving problems
 Thinking abstractly
 Decision-making
 Learning and understanding new
material
 Getting profit from past experience
What is Artificial intelligence
 Designing
computer programs to make
computers smarter
 Ability of an artifact to perform the
same kinds of functions that
characterize human thought
 Machine doing what would be
considered intelligent if done by human
Definitions by categories
Systems that think System that think
like
rationally
humans
Systems that act
like
humans
Systems that act
rationally
Artificial intelligence in a vision
Branches of AI

Logical AI Search
 Search
 Pattern Recognition
 Representation
 Inference
 Common Sense Knowledge and Reasoning
 Virtual reality
What are Neural Nets?

A neural net is an artificial representation of
the human brain that tries to simulate its
learning process.

The term "artificial" means that neural nets
are implemented in computer programs that
are able to handle the large number of
necessary calculations during the learning
process
The biological model
 The
human brain consists of a large
number (more than a billion) of neural
cells that process information's.
 Each cell works like a simple processor
and only the massive interaction
between all cells and their parallel
processing makes the brain's abilities
possible
Biological Model
Artificial Neuron
AI Pioneers
 Alan
M. Turing
 “Computing
 Marvin
Machinery and Intelligence”
Minksy
 Constructed
 Herbert
the first neural net machine
Simon, Allen Newell, J.C. Shaw
 Developed
the first AI computer program
Famous AI Programs

ELIZA (Joseph Weizenbaum)
 Psychologist
 Deep Blue (IBM)
 Chess program
 Cyc (MCC and Cycorp)


Multi-contextual knowledge base and inference
engine
HAL (Arthur C. Clarke)

Space explorer
Success or applications of AI
Face Recognition
 Bacterial Diagnosis
 Mass Spectroscopy
 Speech dictation
 IQ tests
 Genetic
 Probability
 DNA/Protein
 Chess

Success or applications of AI
Driving a car
 Weather
 Sports
 Optimizing Nuclear Power
 Automatic fuzzy clustering
 E-design (bridges, houses)
 E-government decision and planning
 E-commerce
 E-learning
 Robotics
 Translation

Limitation of AI

Cognitive sciences still have not succeeded
in determining exactly what the human
abilities are.
 program designers lack understanding of the
intellectual mechanisms required to do the
task efficiently as that of expert person.
 One is biological so AI should study more
humans and imitate psychology and
physiology.
The Turing Test
 Motivated
to identify intelligence in a
computer program.
 Proposed in 1950 by Alan Turing.
 Original Proposal:
 Given
a person X, a computer Y, and an
interrogator C, C isolated from X and Y.
 C must determine who is the person
 X is intelligent if it can fool C.
Person
Interrogator
Computer
Problems with the Turing Test
 Intelligence
may be considered as a
continuum. The Turing test only
identifies one (very strong) type of
intelligence, and thus offers no means
to measure.
 Does fooling C really imply intelligence?
Our Proposal
 Motivated



to allow:
a measure of intelligence.
more rigid definitions.
more flexible admission of programs.
Example
 The
set of all Math Problems is a
maximal element.
 If a program can solve these problems,
it is said to be an Expert in Math
Problems.
Machine Learning
What Is Machine Learning?
 Enabling
machines to process data in
such a way that it can be to make future
decisions
 ML been studied for many years
 ML has many applications in a variety of
fields
Tasks For Machines
 Pattern
recognition
 Grouping/classification
 Strategizing
 Generating heuristics
 Problem solving
Issues in Machine Learning
 Computational
complexity
 Ethics
 Correctness
 Would
the exact desired learning be
constructed?
 What if there is an error in learning?
Chess & AI
Chess
 In
the early 1950's Alan Turing wrote the
first modern chess program
 Chess was one of the original AI
problems, because it represented one
form of intelligent behavior
 Most successful chess programs use
some kind of search algorithm as their
foundation for choosing move
Chess Algorithms

Chess base on search Algorithm
 Each piece can be assigned a number
indicating its rank (pawns are worth 1, knights
and bishops 3, rooks 5, queens 9 )
 This figure can be multiplied by another
number indicating the strength of the piece's
position on the board
 Computer tries to choose the move that will
leave it in the strongest position
Do this by evaluation function
Chess Algorithms
 Other
formulas quantify concepts like
"king safety," or how well protected that
piece is. These rules, called heuristics,
but they give the computer a rough
sense of the state of the game and a
basis on which to make its decisions
AI based movies
Matrix
In the near future, a computer hacker
named Noe discovers that all life on Earth
may be nothing more than an elaborate
illusion created by a malevolent cyberintelligence, for the purpose of placating us
while our life essence is “farmed” to the
Matrix’s campaign
Matrix Reloaded
Neo, Morpheus, Trinity, and the rest pf
their crew continue to battle the
Machine that have enslaved the human
race in the matrix. Now, more human
are waking up out of the matrix and
attempting to live
Matrix Revolution
In the third installment, the epic war
between man and machine reaches a
thundering crescendo: the Zion military
aided by courageous civilian volunteers
like Zee and the kid desperately battles
to hold back the Sentinel invasion as
the machine army bores into their
stronghold
X–Files Episode7
A computer with highly advanced
Artificial Intelligence begins to kill
in order to preserve its existence
when it is deemed inefficient to
continue controlling the workings
of an office building.
Rob La Belle programmer
of AI Computer
Future of Artificial Intelligence
However, there is little doubt among the
community that artificial machines will be
capable of intelligent thought in the near
future. It's just a question of what and when...
The machines may be pure silicon, quantum
computers or hybrid combinations of
manufactured components and neural tissue.
As for the date, expect great things to happen
within this century
Thanks for your Cooperation