Artificial intelligence

From ArticleWorld


Artificial Intelligence  : This term was coined by John McCarthy, the inventor of Lisp, in 1956. It falls under the area of computer science and it deals with intelligent behavior, machine learning and adaptation.

Researchers in this area seek to automate tasks. While computers can be programmed to make decisions using decision trees and control structures, making an intelligent choice given a variety of circumstances is a human skill. To equip machines to do this, the researchers have to deal with machine learning and adaptation, to name a few areas.


[edit]

AI in daily life

Control, scheduling, planning, diagnosis, answering consumer questions are all examples of the use of AI. Handwriting, facial anmd speech recognition are also the slighly more complex tasks automated using this new discipline.

It is an excellent example of a new discipline arising as a result of scientists looking to reduce tedium. Today it is used in almost every field. It is used extensively in forecasting techniques and defense scientists, not to mention economics, medicine engineering. Many household appliances use chips equipped with AI software to adapt to the needs of the user.

[edit]

Classification

There are two schools of thought in AI: Conventional or Good Old fashioned AI (GOFAI) Computational

Now known as machine learning, the conventional kind is characterized by statistical analysis of the data gathered and formalism. (Other names: symbolic AI, logical AI, and neat AI)

Four main procedures are involved

  • Expert systems
  • Case based reasoning
  • Bayesian networks
  • Behavior based AI

An expert system can process large amounts and complex information to provide conclusions based on them. Computational intelligence on the other hand consists of learning based on empirical data.

Three systems are recognized.

  • Systems with very strong pattern recognition capabilities are based on neural networks. The reasoning behind building a neural network is that simple neurons can be connected to forma network with great processing power.
  • Fuzzy systems are based in fuzzy logic which is the alternative to binary logic, that is, systems with zeros and ones for out comes. Since there is some uncertainity involved, these systems are known as fyzzy systems. They find a use in modern industrial and consumer product control systems.
  • The third system is a little more complex because it’s a set of biological concepts adapted by computer scientists to come up with the most effective algorithm. The biological concepts involved are evolution, populations, mutations and finally survival of the fittest algorithm.

Systems based in biological concepts are, further, of two kinds:

  • Evolutionary algorithms
  • Swarm intelligence

The latest approach that covers all bases is a hybrid approach.

[edit]

A brief history of AI

René Descartes expressed that he thought of animals as nothing more than complex machines as early as the 17th century. The first mechanical digital calculating machine was invented and used in 1642 by Pascal (of Pascal’s Law). Bertrand Russell co-authored Principia Mathematica, revolutionizing formal logic.


  • 1951  : The first AI programs a draughts-game software created by

Christopher Strachey and a chess-game software created by Dietrich Prinz.

  • 1960s - 1970s  : The first knowledge-based program in mathematics created by J.Moses.
  • 1980s  : Widespread usage of ANNs.
  • 1990s  : Deep Blue, a chess-playing computer.
Retrieved from "http://www.articleworld.org/index.php/Artificial_intelligence"