ABSTRACT Such programs should be considered ``somewhat intelligent. It is related to the This paper is the introduction to similar task of using computers toArtificial intelligence (AI). Artificial understand human intelligence.intelligence is exhibited by artificial entity, a We can learn something about howsystem is generally assumed to be a to make machines solve problems bycomputer. AI systems are now in routine use observing other people or just by observingin economics, medicine, engineering and the our own methods. On the other hand, mostmilitary, as well as being built into many work in AI involves studying the problemscommon home computer software the world presents to intelligence rather thanapplications, traditional strategy games like studying people or animals. AI researcherscomputer chess and other video games. are free to use methods that are not observed We tried to explain the brief ideas of in people or that involve much moreAI and its application to various fields. It computing than people can do. Wecleared the concept of computational and discussed conditions for considering aconventional categories. It includes various machine to be intelligent. We argued that ifadvanced systems such as Neural Network, the machine could successfully pretend to beFuzzy Systems and Evolutionary human to a knowledgeable observer thencomputation. AI is used in typical problems you certainly should consider it intelligent.such as Pattern recognition, Naturallanguage processing and more. This system INTRODUCTION:-is working throughout the world as an Artificial intelligence (AI) :-artificial brain. Artificial intelligence (AI) is defined Intelligence involves mechanisms, as intelligence exhibited by an artificialand AI research has discovered how to make entity. Such a system is generally assumedcomputers carry out some of them and not to be a computer.others. If doing a task requires only Although AI has a strong sciencemechanisms that are well understood today, fiction connotation, it forms a vital branch ofcomputer programs can give very computer science, dealing with intelligentimpressive performances on these tasks. behaviour, learning and adaptation in
machines. Research in AI is concerned with system can process large amounts of known producing machines to automate tasks information and provide conclusions based requiring intelligent behavior. Examples on them. include control, planning and scheduling, • Case based reasoning the ability to answer diagnostic and • Bayesian networks consumer questions, handwriting, speech, • Behavior based AI: a modular and facial recognition. As such, it has method of building AI systems by hand. become a scientific discipline, focused on Computational Intelligence (CI) :- providing solutions to real life problems. AI Computational Intelligence involves systems are now in routine use in iterative development or learning (e.g. economics, medicine, engineering and the parameter tuning e.g. in connectionist military, as well as being built into many systems). Learning is based on empirical common home computer software data and is associated with non-symbolic AI, applications, traditional strategy games like scruffy AI and soft computing. computer chess and other video games. Categories of AI :- Methods include: AI divides roughly into two schools of • Neural networks: systems with very thought: strong pattern recognition• Conventional AI. capabilities.• Computational Intelligence (CI). • Fuzzy systems: techniques for Conventional AI :- reasoning under uncertainty, has been Conventional AI mostly involves widely used in modern industrial and methods now classified as machine learning, consumer product control systems. characterized by formalism and statistical • Evolutionary computation: applies analysis. This is also known as symbolic AI, biologically inspired concepts such as logical AI, neat AI and Good Old Fashioned populations, mutation and survival of the Artificial Intelligence (GOFAI). fittest to generate increasingly better Methods include: solutions to the problem. These methods• Expert systems: apply reasoning most notably divide into evolutionary capabilities to reach a conclusion. An expert
algorithms (e.g. genetic algorithms) and must include the ability to adapt to newswarm intelligence (e.g. ant algorithms). circumstances. Psychologists generally do not characterize human intelligence by just one trait but by the combination of many diverse abilities. Research in AI has focused chiefly on the following components of intelligence: learning, reasoning, problem solving, perception, and using language. Problem solving Problem solving, particularly in artificial intelligence, may be characterized as a systematic search through a range of possible actions in order to reach some predefined goal or solution. Problem-solving methods divide into special purpose and general purpose. A special-purpose method is tailor-made for a particular problem andWhat is intelligence? often exploits very specific features of the All but the simplest human situation in which the problem is embedded.behaviour is ascribed to intelligence, while In contrast, a general-purpose method iseven the most complicated insect behaviour applicable to a wide variety of problems.is never taken as an indication of One general-purpose technique used in AI isintelligence. What is the difference? means-end analysis—a step-by-step, orConsider the behaviour of the digger wasp, incremental, reduction of the differenceSphex ichneumoneus. When the female between the current state and the final goal.wasp returns to her burrow with food, she The program selects actions from a list offirst deposits it on the threshold, checks for means—in the case of a simple robot thisintruders inside her burrow, and only then, if might consist of PICKUP, PUTDOWN,the coast is clear, carries her food inside. MOVEFORWARD, MOVEBACK,The real nature of the wasps instinctual MOVELEFT, and MOVERIGHT—until thebehaviour is revealed if the food is moved a goal is reached.few inches away from the entrance to her Many diverse problems have beenburrow while she is inside: on emerging, she solved by artificial intelligence programs.will repeat the whole procedure as often as Some examples are finding the winningthe food is displaced. Intelligence— move (or sequence of moves) in a boardconspicuously absent in the case of Sphex—
game, devising mathematical proofs, and eye and a pincer hand, constructed at themanipulating “virtual objects” in a University of Edinburgh, Scotland, duringcomputer-generated world. the period 1966–73 under the direction of Donald Michie. FREDDY was able toPerception recognize a variety of objects and could be instructed to assemble simple artifacts, such as a toy car, from a random heap of components. APPLICATIONS OF AI :- Game Playing :- You can buy machines that can play master level chess for a few hundred dollars. There is some AI in them, but they play well against people mainly through brute force computation--looking at hundreds of thousands of positions. In perception the environment isscanned by means of various sensory Speech Recognition :-organs, real or artificial, and the scene isdecomposed into separate objects in various In the 1990s, computer speech recognitionspatial relationships. Analysis is reached a practical level for limitedcomplicated by the fact that an object may purposes. Thus United Airlines has replacedappear different depending on the anglefrom which it is viewed, the direction and its keyboard tree for flight information by aintensity of illumination in the scene, and system using speech recognition of flighthow much the object contrasts with the numbers and city names. It is quitesurrounding field. convenient. On the other hand, while it is At present, artificial perception is possible to instruct some computers usingsufficiently well advanced to enable optical speech, most users have gone back to thesensors to identify individuals, autonomousvehicles to drive at moderate speeds on the keyboard and the mouse as still moreopen road, and robots to roam through convenient.buildings collecting empty soda cans. Oneof the earliest systems to integrate Understanding Natural Language :-perception and action was FREDDY, a Just getting a sequence of words intostationary robot with a moving television a computer is not enough. Parsing sentences
is not enough either. The computer has to be than medical students or practicing doctors, provided with an understanding of the provided its limitations were observed. domain the text is about, and this is Heuristic Classification :- presently possible only for very limited One of the most feasible kinds of domains. expert system given the present knowledge Computer Vision :- of AI is to put some information in one of a The world is composed of three- fixed set of categories using several sources dimensional objects, but the inputs to the of information. An example is advising human eye and computer’s TV cameras are whether to accept a proposed credit card two dimensional. Some useful programs can purchase. Information is available about the work solely in two dimensions, but full owner of the credit card, his record of computer vision requires partial three- payment and also about the item he is dimensional information that is not just a set buying and about the establishment from of two-dimensional views. At present there which he is buying it (e.g., about whether are only limited ways of representing three- there have been previous credit card frauds dimensional information directly, and they at this establishme are not as good as what humans evidently Conclusion:- use. We conclude that if the machine Expert Systems :- could successfully pretend to be human to a A ``knowledge engineer interviews knowledgeable observer then you certainly experts in a certain domain and tries to should consider it intelligent. AI systems are embody their knowledge in a computer now in routine use in various field such as program for carrying out some task. How economics, medicine, engineering and the well this works depends on whether the military, as well as being built into many intellectual mechanisms required for the task common home computer software are within the present state of AI. One of the applications, traditional strategy games etc. first expert systems was MYCIN in 1974, which diagnosed bacterial infections of the AI is an exciting and rewarding blood and suggested treatments. It did better discipline. AI is branch of computer science that is concerned with the
automation of intelligent behavior. The revised definition of AI is - AI is the study of mechanisms underlying intelligent behavior through the construction and evaluation of artifacts that attempt to enact those mechanisms. So it is concluded that it work as an artificial human brain which have an unbelievable artificial thinking power. Programs with Common Sense :- John McCarthy, In Mechanization of Thought Processes, Proceedings of the Symposium of the National Physics Laboratory, 1959. Artificial Intelligence, Logic and Formalizing Common Sense: - Richmond Thomason, editor, Philosophical Logic and Artificial Intelligence. Klüver Academic, 1989.