Artificial intelligence

ARTIFICIAL INTELLIGENCE


COMPILED BY

AjAY MALALIkAR

ANIRuDh DEshPANDE

Department of Computer Science and Engineering Page 1


Abstract and the inference engine or algorithm which uses
the knowledge base to arrive at a conclusion.
It is the science and engineering of making Neural Networks:
intelligent machines, especially intelligent
computer programs. It is related to the similar task NNs are biologically inspired systems consisting of
of using computers to understand human a massively connected network of computational
intelligence, but AI does not have to confine itself “neurons,” organized in layers. By adjusting the
to methods that are biologically observable. While weights of the network, NNs can be “trained” to
no consensual definition of Artificial Intelligence approximate virtually any nonlinear function to a
(AI) exists, AI is broadly characterized as the study required degree of accuracy. NNs typically are
of computations that allow for perception, reason provided with a set of input and output exemplars.
and action. This paper examines features of A learning algorithm (such as back propagation)
artificial Intelligence, introduction, definitions of would then be used to adjust the weights in the
AI, history, applications, growth and achievements. network so that the network would give the desired
1.0 Introduction output, in a type of learning commonly called
supervised learning.

FUZZY SYSTEMS
Fuzzy set theory was proposed by Zadeh (1965) as
a way to deal with the ambiguity associated with
almost all real-world problems. Fuzzy set
membership functions provide a way to show that
an object can partially belong to a group. Classic
Artificial Intelligence (AI) is the set theory defines sharp boundaries between sets,
branch of computer science which deals with which mean that an object can only be a member or
intelligence of machines where an intelligent agent a nonmember of a given set. Fuzzy membership
is a system that perceives its environment and takes functions allow for gradual transitions between sets
actions which maximize its chances of success. It is and varying degrees of membership for objects
the study of ideas which enable computers to do the within sets. Complete membership in a fuzzy
things that make people seem intelligent. The function is indicated by a value of +1, while
central principles of AI include such traits as complete non-membership is shown by a value of
reasoning, knowledge, planning, learning, 0. Partial membership is represented by a value
communication, perception and the ability to move between 0 and +1.
and manipulate objects. It is the science and
engineering of making intelligent machines,
especially intelligent computer programs 2.0 Some definitions of AI
1.1 ARTIFICIAL INTELLIGENCE METHODS:
At the present time, AI methods can be divided into  Computers with the ability to mimic or
two broad categories: (a) symbolic AI, which duplicate the functions of the human brain
focuses on the development of knowledge-based
systems (KBS); and (b) computational intelligence,  Artificial Intelligence (AI) is the study of
which includes such methods as neural networks how computer systems can simulate
(NN), fuzzy systems (FS), and evolutionary intelligent processes such as learning,
computing. A very brief introduction to these AI reasoning, and understanding symbolic
methods is given below, and each method is information in context. AI is inherently a
discussed in more detail in the different sections of multi-disciplinary field. Although it is
this circular. most commonly viewed as a subfield of
computer science, and draws upon work in
1.1 Knowledge-Based Systems: algorithms, databases, and theoretical
computer science, AI also has close
A KBS can be defined as a computer system connections to the neurosciences,
capable of giving advice in a particular domain, cognitive science and cognitive
utilizing knowledge provided by a human expert. A psychology, mathematical logic, and
distinguishing feature of KBS lies in the separation engineering."
behind the knowledge, which can be represented in
a number of ways such as rules, frames, or cases,



 ``The exciting new effort to make 3.0 History
computers think ... machines with minds,
in the full and literal sense'' (Haugeland,
The modern history of AI can be traced back to the
1985)
year 1956 when John McCarthy proposed the term
 ``The automation of activities that we as the topic for a conference held at Dartmouth
associate with human thinking, activities College in New Hampshire devoted to the subject.
such as decision-making, problem solving, The initial goals for the field were too ambitious
learning ...'' (Bellman, 1978) ``The and the first few AI systems failed to deliver what
study of mental faculties through the use was promised. After a few of these early failures,
of computational models'' (Charniak and AI researchers started setting some more realistic
McDermott, 1985) goals for themselves. In the 1960s and the 1970s,
the focus of AI research was primarily on the
 ``The study of the computations that make development of KBS or expert systems. During
it possible to perceive, reason, and act'' these years, expert systems technology were
(Winston, 1992) applied to a wide range of problems and fields
ranging from medical diagnosis to inferring
 ``The art of creating machines that molecular structure to natural language
perform functions that require intelligence understanding. The same period also witnessed
when performed by people'' (Kurzweil, early work on NNs, which showed how a
1990) distributed structure of elements could collectively
represent an individual concept, with the added
 ``The study of how to make computers do advantage of robustness and parallelism. However,
things at which, at the moment, people are the publication of Minsky and Papert’s book
better'' (Rich and Knight, 1991) ``A Perceptrons in 1969, which argued for the limited
field of study that seeks to explain and representation capabilities of NN, led to the demise
emulate intelligent behavior in terms of of NN research in the 1970s.
computational processes'' (Schalkoff,
1990) The late 1980s and the 1990s saw a renewed
interest in NN research when several different
 ``The branch of computer science that is researchers reinvented the back propagation
concerned with the automation of learning algorithm (although the algorithm was
intelligent behavior'' (Luger and really first discovered in 1969). The back
Stubblefield, 1993) propagation algorithm was soon applied to many
learning problems causing great excitement within
 "Artificial intelligence is the study of the AI community. The 1990s also witnessed some
ideas to bring into being machines that dramatic changes in the content and methodology
respond to stimulation consistent with of AI research. The focus of the field has been
traditional responses from humans, given shifting toward grounding AI methods on a
the human capacity for contemplation, rigorous mathematical foundation, as well as to
judgment and intention. Each such tackle real-world problems and not just toy
machine should engage in critical examples. There is also a move toward the
appraisal and selection of differing development of hybrid intelligent systems (i.e.,
opinions within itself. Produced by human systems that use more than one AI method)
skill and labor, these machines should stemming from the recognition that many AI
conduct themselves in agreement with methods are complementary. Hybrid intelligent
life, spirit and sensitivity, though in systems also started to use newer paradigms that
reality, they are imitations." mimic biological behavior such as GAs and fuzzy
logic.
 It is the science and engineering of
making intelligent machines, especially
intelligent computer programs. It is related 4.0 Applications of AI
to the similar task of using computers to
understand human intelligence, but AI
does not have to confine itself to methods 4.1 Finance:
that are biologically observable.
Banks use artificial intelligence systems to
organize operations, invest in stocks, and manage
properties. In August 2001, robots beat humans in a
simulated financial trading competition. Financial
institutions have long used artificial neural network



systems to detect charges or claims outside of the discrepancies between expected and actual
norm, flagging these for human investigation. operating conditions.
Some other applications in this section include loan
investigation, ATM design, safe and fast banking SCHEDULING AND PLANNING: In the present
etc. day world TIME ELEMENT plays an important
role. So, completion of manufacturing within short
4.2 Medicine: period of time in addition to good quality becomes
very important. Intelligent software offers several
A medical clinic can use artificial intelligence advantages in developing computerized scheduling
systems to organize bed schedules, make a staff systems. Instead of presenting one optimization
rotation, and provide medical information. schedule, AI-based scheduling systems present
Artificial neural networks are used for medical several schedules with their evaluation indexes.
diagnosis, functioning as Machine differential The operator can then select the "best" optimum
diagnosis. AI has also application in fields of schedule.
cardiology (CRG), neurology (MRI), embryology
(sonography), complex operations of internal 4.4 Telecommunications:
organs etc.
Many telecommunications companies make use of
4.3 Heavy Industry: heuristic search in the management of their
workforces, for example BT Group has deployed
Now a days in big industries all the work and heuristic search in a scheduling application that
machine operations are controlled by principles of provides the work schedules of 20000 engineers.
Artificial Intelligence. These huge machines
involve risk in their manual maintenance and 4.5 Music:
working. So in becomes necessary part to have an
efficient and safe operation agent in their operation. AI, scientists are trying to make the computer
emulate the activities of the skillful musician.
4.3.1 Application Types and Situations: Composition, performance, music theory, sound
Intelligent software systems play a number of roles processing are some of the major areas on which
in heavy industry. Selected examples are discussed research in Music and Artificial Intelligence are
below. focusing on. Some of them are

PROCESS CONTROL: These tasks usually OrchExtra: This program was designed to provide
involve automation of low-level control in a real- small-budget productions with instrumentation for
time system. The implemented systems are all instruments usually present in the full-fledged
concerned with fault detection, diagnosis and orchestra.
alarming, and with operating the control devices in
the control loops. Integral functions of intelligent COMPUTER ACCOMPNIMENT: The Computer
software are sensor diagnostics, handling of Music Project at CMU develops computer music
erroneous or missing data, and performing and interactive performance technology to enhance
temporal reasoning. human musical experience and creativity.

PROCESS MONITORING: Artificial Intelligence SMART MUSIC: SmartMusic is an interactive,
systems monitor, compare and analyze the process computer-based practice tool for musicians.
behavior of events that are crucial to successful
operation and suggest any corrective action that CHUCK: Chuck is a text-based, cross-platform
should be implemented by the operators. language that allows real-time synthesis,
composition, performance and analysis of music.
FAULT DIAGNOSIS AND MAINTAINANCE: It
is practically impossible to diagnose huge 4.6 Antivirus:
machines regularly and precisely. Working of
faulty machines may cause great loss to the Artificial intelligence (AI) techniques have played
industry. So, Artificial Intelligence systems offer a increasingly important role in antivirus detection.
number of advantages for working with diagnostic At present, some principal artificial intelligence
problems. First, they can monitor and analyze techniques applied in antivirus detection are
hundreds of sensors, determine any anomalies in proposed, including heuristic technique, data
their functions and identify probable causes of the mining, agent technique, artificial immune, and
artificial neural network. It improves the



performance of antivirus detection systems, and of robot software is to create robotic applications
promotes the production of new artificial that help or entertain people. Applications include
intelligence algorithm and the application in command-and-control and tasking software.
antivirus detection to integrate antivirus detection Command-and-control software includes robot
with artificial intelligence. This paper introduces control GUIs for tele-operated robots, point-n-click
the main artificial intelligence technologies, which command software for autonomous robots, and
have been applied in antivirus system. Meanwhile, scheduling software for mobile robots in factories.
it also points out a fact that combining all kinds of Tasking software includes simple drag-n-drop
artificial intelligence technologies will become the interfaces for setting up delivery routes, security
main development trend in the field of antivirus. patrols and visitor tours; it also includes custom
programs written to deploy specific applications.
General purpose robot application software is
deployed on widely distributed robotic platforms.
4.7 Robotics:

4.8 Gaming:

Definition: What is a Robot?

Robots are physical agents that perform tasks by
manipulating the physical world. They are In the earlier days gaming technology was not
equipped with sensors to perceive their broadened. Physicist Willy Higinbotham created
environment and effectors to assert physical forces the first video game in 1958. It was called “Tennis
on it (covered in more detail in next section). For Two” and was plan oscilloscope. But now AI
Robots can be put into three main categories: technology has become vast and standard has also
manipulators, mobile robots and humanoid. been increased. More realistic, heavily graphical, 3-
D games are been designed by developers. Some of
Robotics and AI most popular games of present day are Crisis, Fear,
Artificial intelligence is a theory. The base object is Fall Out, Halo etc.
the agent who is the "actor". It is realized in
software. Robots are manufactured as hardware.
The connection between those two is that the 4.9 SOME OTHER APPLICATIONS:
control of the robot is a software agent that reads
data from the sensors decides what to do next and  Credit granting
then directs the effectors to act in the physical
world.  Information management and retrieval

 AI and expert systems embedded in
Robot application software products

Most robot  Plant layout
manufacturers keep
their software  Help desks and assistance
hidden. It is
impossible to find  Employee performance evaluation
out how most robots
are programmed. It  Shipping
is almost as if they
 Marketing
had no software in
many cases.  Warehouse optimization
Regardless which language is used, the end result



 In space workstation maintenance
 Satellite controls

 Network developments

 Military activity controls

 Nuclear management

5.0 The Explosive Growth of AI

 Deep Thought is an international
Since AI is applicable in almost all fields, they grand master chess player.
become the needs of our life .It make the
development of AI field. It is the reason behind the  Sphinx can recognize continuous
explosive growth of AI. speech without training for each
speaker. It operates in near real
The growth can be divided into two parts based on time using a vocabulary of 1000
the application area and what purpose the used, words and has 94% word
they are as follows accuracy.
1. Growth in positive sense(useful to society)  Navlab is a truck that can drive
along a road at 55mph in normal
2. Growth in negative sense(harmful to
traffic.
society)
 Carlton and United Breweries use
an AI planning system to plan
production of their beer.

 Natural language interfaces to
databases can be obtained on a
PC.

6.0 Some Achievements of AI  Machine Learning methods have
been used to build expert
systems.
 DARPA Grand challenge- 123
 Expert systems are used regularly
miles through the desert
in finance, medicine,
manufacturing, and agriculture

7.0 Future of AI

Having discussed about AI one debatable question
arises that is artificial intelligence more powerful
than natural intelligence. Looking at the features
and its wide applications we may definitely stick to
artificial intelligence. Seeing at the development of
AI, is it that the future world is becoming artificial.

 DARPA Urban Challenge- Biological intelligence is fixed, because it
Autonomous driving in traffic is an old, mature paradigm, but the new paradigm
of non-biological computation and intelligence is
growing exponentially. The crossover will be in the



2020s and after that, at least from a hardware • Gardner, Martin. Logic Machines &
perspective, non-biological computation will Diagrams. NY: McGraw-Hill, 1958.
dominate…

The memory capacity of the human brain
is probably of the order of ten thousand million
binary digits. But most of this is probably used in
remembering visual impressions, and other
comparatively wasteful ways. One might
reasonably hope to be able to make some real
progress [towards artificial intelligence] with a few
million digits [of computer memory].

Hence we can say that as natural
intelligence is limited and volatile too world may
now depend upon computers for smooth working.

8.0 Conclusion

Till now we have discussed in brief about
Artificial Intelligence. We have discussed
some of its principles, its applications, its
achievements etc. The ultimate goal of
institutions and scientists working on AI is
to solve majority of the problems or to
achieve the tasks which we humans
directly can’t accomplish. It is for sure
that development in this field of computer
science will change the complete scenario
of the world. Now it is the responsibility
of creamy layer of engineers to develop
this field.

9.0 References

http://en.wikibooks.org/wiki/Computer_Science:
Artificial_Intelligence
http://www.howstuffworks.com/arificialintellige
nce
http://www.google.co.in
http://www.library.thinkquest.org

• Cohen, Jonathan. Human Robots in Myth
and Science. NY: A.S.Barnes, 1967.

• Feigenbaum, E.A. & Feldman, J. (eds.)
Computers and Thought. NY: McGraw-
Hill, 1963.


Artificial intelligence

More Related Content

What's hot (20)

Similar to Artificial intelligence (20)

Artificial intelligence