Sunteți pe pagina 1din 40

Man v/s Computer

Introduction

The term Artificial Intelligence was
coined by-John McCarthy in 1956.
Artificial Intelligence is concerned with the
design of intelligence in an artificial device.
Intelligence.
Artificial Device.
What Intelligence?

Is it that which characterize humans? Or is there
an absolute standard of judgement?
Two possibilities:
System Intelligence = human intelligence
System Should behave in best manner
Secondly what type of behaviour are we talking
about?
Are we looking at the thought process or reasoning ability of the system?
OR
Final manifestations of the system in terms of its actions?

Ingredients for Intelligence..

.

Experience
Knowledge
Imagination
Creativity
Reaction to
situation
Visualization
Manipulation
Intelligence
So what is A.I then?

Scopes and views on A.I:
I. Designing systems that are as intelligent
as humans :understanding human
thoughtto build similar machines.
II. Based on Turing test

Are you Man or Machine ??
Turing Test:

E.g. CAPTCHA

-prevents flooding websites

The third view of AI is that it is the study of
rational agents; building machines that act
rationally. The focus is on how the system
acts and performs, and not so much on
the reasoning process.
A rational agent is one that acts rationally,
that is, is in the best possible manner.

Typical A.I. problems

Tasks include common place and expert tasks
Examples of common-place tasks include
Recognizing people, objects.
Communicating (through natural language).
Navigating around obstacles on the streets

Examples of expert tasks
Medical diagnosis.
Mathematical problem solving
Playing games like chess



Categories..
Strong AI : True reasoning and problem
solving. Self awareness.
Weak AI: Act ,simulate human cognition.
Applied AI: Development of smart
systems for e.g. a security system.
Cognitive AI: Test theories of hoe the
human mind works; how we recognize
faces.

Classicalism And
Connectionism
Classicalism
Chess,
weather forecast e.t.c
Human minds are
computational by
nature
-Uses a top down
approach
Connectionism

human minds can
learn, expand and
change

-e.g. Neural networks
adaptive algorithms
Areas of A.I.
ROBOTICS
Mechanical devices
that perform tasks
requiring high degree
of precision or that
are hazardous to
humans.
Rover
Uses Genetic Algorithms
(is an A.I Concept)



Genetic Algorithms
Search technique used in computing to find true or approximate solutions to
optimization and search problems

A typical genetic algorithm requires:
a genetic representation of the solution domain,
a fitness function to evaluate the solution domain.
Applied to electronic circuit design : Evolvable hardware ,code breaking and
deciphering.
Genetic algorithms are a part of evolutionary computing, which is a rapidly growing area of artificial intelligence.

The Algorithm at a glance..

[Start] Generate random population of n chromosomes (suitable solutions for the problem)
[Fitness] Evaluate the fitness f(x) of each chromosome x in the population
[New population] Create a new population by repeating following steps until the new population
is complete
[Selection] Select two parent chromosomes from a population according to their fitness (the
better fitness, the bigger chance to be selected)
[Crossover] With a crossover probability cross over the parents to form a new offspring
(children). If no crossover was performed, offspring is an exact copy of parents.
[Mutation] With a mutation probability mutate new offspring at each locus (position in
chromosome).
[Accepting] Place new offspring in a new population
[Replace] Use new generated population for a further run of algorithm
[Test] If the end condition is satisfied, stop, and return the best solution in current population
[Loop] Go to step 2

Overview
A typical chromosome may look like this!
100101011101010010100111011011101111111
01 (encoding a potential solution)

What is crossover?
e.g. Given two chromosomes
10001001110010010
01010001001000011
Choose a random bit along the length, say at
position 9, and swap all the bits after that
point
so the above become:
10001001101000011
01010001010010010




What's the Mutation Rate?

chance that a bit within a chromosome could
beflipped be flipped (0 becomes 1, 1
becomes 0). This is usually a very low value
for binary encoded genes, say 0.001

So whenever chromosomes are chosen from the
population the algorithm first checks to see if
crossover should be applied and then the
algorithm iterates down the length of each
chromosome mutating the bits if applicable.

examples
Given the digits 0 through 9 and the operators +,
-, * and /, find a sequence that will represent
a given target number. The operators will be
applied sequentially from left to right as you
read.
E.g. target number 23, the sequence
is 6+5*4/2+1 and many more..
Encoding(as string of bits).
0: 0000
1: 0001
2: 0010
3: 0011
4: 0100
5: 0101
6: 0110
7: 0111
8: 1000
9: 1001
+: 1010
-: 1011
*: 1100
/: 1101



23, ' 6+5*4/2+1' would be represented by nine genes
like so:
0110 1010 0101 1100 0100 1101 0010 1010 0001
6 + 5 * 4 / 2 + 1

-Strung together as:
011010100101110001001101001010100001
Decoding
0010001010101110101101110010 is randomized.

0010 0010 1010 1110 1011 0111 0010
2 2 + n/a - 7 2
So just 2+7 (since a number has to be followed by an
operator)

Deciding on a Fitness Function could be
intriguing!

Deciding on a Fitness Function
Give a higher fitness score for a chromosome coming closer to the solution
If we assume the target number for the remainder of the tutorial is 42, the chromosome mentioned
above

011010100101110001001101001010100001

-has a fitness score of 1/(42-23) or 1/19.

-if a solution is found, a divide by zero error would occur as the fitness would be 1/(42-42).

Down to business: Genetic Algorithm
e.g.Searching for a maximum of the function , where x can take values between 0 and
31.
f(x)=x
2

obviuos soln:

31

Computer Vision Systems
Hardware and
software that permit
computers to capture,
store & manipulate
images.

contd
Pattern and object recognition most widely
used so far: faces, finger-print, vehicles.
Motion Analysis, e.g. estimate velocities at
each point in the image.
Scene re-construction and image restoration:
constructing 3D models, removal of
noise(sensor noise, motion blur)-filter
implementation
AcquisitionPre-processingFeature
extractionHigh level processingdecision
making






Neural Networks(ANN)
Computing systems developed to act like or simulate the
human brain
Only a part can be simulated :that which is required for
performing a specific task for e.g. sense of smell for
particular items like tea.

Fuzzy Logic and the ABS
Mathematical system used to solve
problems that are expected to have many
solutions
-implemented in computers that make yes-
no decisions.
-works with ranges of values

[Terms:FUZZIFICATION - The crisp input has to be converted to a fuzzy number.
DEFUZZIFICATION- It is the process of producing a quantifiable result in fuzzy logic.]

IF brake temperature IS warm AND speed IS not very fast THEN brake pressure IS
slightly decreased.



The Fuzzy-Controller uses two input values: the wheel slip S
B
:
and the wheel acceleration:

-The input variables are transformed into fuzzy variables slip and wheel acceleration by the
fuzzification process. Both variables use seven linguistic values, the slip variable is described by
the terms
Slip ={zero, very small, too small, smaller than optimum, optimum, too large, very large},
and the acceleration by
dvwheel/dt ={negative large, negative medium, negative small, negative few, zero, positive
small, positive large}.
As a result of two fuzzy variables, each of them having seven labels, 49 different conditions are
possible. The rule base is complete that means, all 49 rules are formulated and all 49 conditions
are allowed.
The two fuzzy input values slip and wheel acceleration can be mapped to the fuzzy output value
pressure. The labels for this value are:
Pressure = {positive fast, positive slow, zero, negative slow, negative fast}

Natural Language Processing
Computers react to
statements and
commands made in a
natural language
like English
Agents
Layer 0-avoids obstacles:
sonar scan, generate halt
message when collides.
Layer 1-wander behavior:
generates random heading
Layer 3-Exploration behavior:
notice idle time an looks for an
interesting place. A path-plan
sends new direction to avoid.
Problem solving-the heart of AI
State space search notations
1. Initial state: description of starting
config. Of the agent
2. Action or operator takes the agent
from on state to another(sucessor
states)
3. Plan: sequence of actions.
4. Cost of a plan(positive number):path
cost.



Concept
Problem formulation: choosing a relavant
set of states and a feasible set of
operators for moving from one state to
another.
Search is the process of considering
various possible sequences applied to
the initial state and finding out a
sequence which culminates in a goal
state.

States: nodes
Actions:arcs
Simple Pegs and disks problem

Operators : one may move the top most disk on any needle to the topmost
Position to any other needle.
Initial state Goal state
Problem continues
Sequence of actions:
I. Move AC
II. Move AB
III. Moe AC
IV. Move BA
V. Move CB


Move AB
Move CB
Path Finding
What is Path finding?
How to choose the right path?




e.g. using Dijkstras Algorithm.
-used very much in game development, robot - motion planning.
Dijkstras Algorithm
Introduction:
-created by Edsger Dijkstra in 1959. It advances depending on the cost that each edge
represents, the edges with lower cost will always be chosen first so when you reach
the target, you will be sure that the path is the lowest costing path available.
definitions:
-Nodes : vertices
-Edges : roads
-Graph : Map

basic steps of the algorithm

Take the closest node not yet analyzed.
Add its best edge to the Shortest Path Tree.
If it is the target node, finish the search.
Retrieve all the edges of this node.
For each edge calculate the cost of moving
from the source node to the arrival Node.
If the total cost of this edge is less than the
cost of the arrival node until now, then
update the node cost with the new one.
Take the next closest node not yet analyzed.
Repeat this until the target is reached, or
there are no more nodes available.



LISP AND PROLOG
Simple syntax: mostly
parenthesis and words
Basic elements: s-
expressions
Set of function definition
Not case sensitive
Parenthesis are used to
make list ((+22))
Many special symbols and
square brackets
Terms
Set of facts(info and rules of
inference)-like a relational
database with super powers.
Case sensitive
Parenthesis show
grouping((+22)) is the same
as+22
Todays AI systems
Limited success:
I. Computer vision face recognition
II. Robotics mostly autonomous
III. NLP Simple machine translation

Thank you
artificial intelligence is no match for natural stupidity

S-ar putea să vă placă și