Artificial intelligence (siferrsa)

Artificial intelligence


Artificial Intelligence (AI) or in the Indonesian language is called artificial intelligence. Intelligence, which means bright and artificial is artificial. Artificial intelligence engine which is capable of thinking, weighing the actions taken, and able to make decisions like humans do. Several kinds of fields that use artificial intelligence expert systems, among others, computer games (games), fuzzy logic, neural networks and robotics.
Here are some definitions from the experts on artificial intelligence• H.A.Simon (1897):Artificial intelligence (AI) is an area of ​​research, application and instructions related to programming a computer to do something in the view of man is intelligent.• Rich and Knight (1991):Artificial intelligence (AI) is a study of how to make computers do things which at present can be done better by humans.• Encyclopedia Britannica:Artificial intelligence (AI) is a branch of science in computer science knowledge mempresentasi more use of the form of symbols of the numbers, and process information based on heuristic methods or by virtue of a rule.

Purpose of Artificial Intelligence
According to Winston and Prendergast 1984 goal of artificial intelligence are:
A. Making machines become more intelligent (primary objective)2. Understand what intelligence (scientific objective)3. Make machines more useful (entrepreneurial purpose)

Advantages and Disadvantages of Artificial Intelligence

 Artificial Intelligence advantages: Artificial intelligence is more permanent.Artificial intelligence  easily duplicated and distributed. Artificial intelligence is consistent. Artificial intelligence can be in the documentation. Artificial intelligence can do the job faster than natural intelligence. Artificial intelligence is more efficient. Artificial intelligence can do the job better than natural intelligence.

 Disadvantages of Artificial Intelligence:- Working with the input - the input symbolic.- Artificial intelligence may be used only on a limited basis.- No creative

Schools of thoughtBroadly speaking, the AI ​​is divided into two schools of thought namely Conventional AI and Computational Intelligence (CI, Computational Intelligence). Conventional AI mostly involves methods now diklasifiksikan as machine learning, characterized by formalism and statistical analysis. Also known as symbolic AI, logical AI, AI and AI pure old fashioned way (GOFAI, Good Old Fashioned Artificial Intelligence).

Method-the method include:• Expert systems: the capability to apply judgment to reach conclusions. An expert system can process large amounts of known information and provide conclusions based on such information.• considerations based on case• Bayesian Network• behavior-based AI: a modular method to the formation of AI systems manuallyComputational intelligence involves the development or interactive learning (such as parameter tuning in connectionist systems). Learning is based on empirical data and are associated with non-symbolic AI, AI irregular and soft computing. Basic methods include:• Neural networks: systems with pattern recognition capabilities are very strong.• Fuzzy systems: techniques for consideration under uncertainty, has been used extensively in modern industrial and consumer product control systems.• Evolutionary Computation: applying concepts such as biologically inspired population, mutation and survival of the fittest to produce better solutions.These methods are mainly divided into evolutionary algorithms (eg genetic algorithms) and swarm intelligence (eg ant algorithms).With hybrid intelligent systems, experiments designed to combine these two groups. Expert inference rules can be generated through a neural network or production rules from statistical learning such as the ACT-R. A promising new approach is mentioned that the strengthening of intelligence to try to achieve artificial intelligence in the process of evolutionary development as a side effect of the strengthening of human intelligence through technology.


Soft Computing
 Soft computing is an innovation in building intelligent systems are systems that have expertise in certain domains such as human beings, able to adapt and learn to work better if the environment changes. Soft computing exploits the tolerance for imprecision, uncertainty, and partial truth to be completed and controlled with ease to match reality (Prof. Lotfi Zadeh A, 1992).Methodologies used in Soft computing is:A. Fuzzy Logic / Fuzzy Logic (to accommodate the inaccuracies).2. Artificial Neural Networks / Neurall Network (using learning).3. Probabilistic Reasoning (accommodates uncertainty).4. Genetic Algorithm / Evolutionary Computing (optimization)
CONCEPT OF ARTIFICIAL INTELLIGENCE
A. Game (Game) Most of the game is done using a set of rules. In a game to use what is called the search space. alternative technique to determine the listening problems of space is something that complicated. The technique is called a heuristic. The game is an exciting field in the study of heuristic.
2. NATURAL LANGUAGEA technology that provides the ability for computers to understand human language so that computer users can communicate with computers using everyday language.
3. Robotics and SENSOR SYSTEMSensor systems, such as vision systems, tactile systems, and signal processing systems, when combined with AI, can be categorized into a broad system called robotic systems.
4. EXPERT SYSTEMExpert system (Expert System) is a computer-based advisory programs that try to mimic the thought processes and knowledge of an expert in solving specific problems.

History of Artificial Intelligence
In the early 17th century, René Descartes proposed that bodies of animals are nothing but just complicated machines. Blaise Pascal invents the first mechanical digital calculating machine in 1642. At 19, Charles Babbage and Ada Lovelace worked on mechanical calculating machines can be programmed.Bertrand Russell and Alfred North Whitehead published Principia Mathematica, which revolutionized formal logic. Warren McCulloch and Walter Pitts published "Logical Calculus of Ideas that remain in activity" in 1943 which laid the foundation for neural networks.The 1950s were a period of active efforts in AI. The first working AI programs were written in 1951 to run the Ferranti Mark I machine at the University of Manchester (UK): a game program script written by Christopher Strachey and a chess program written by Dietrich Prinz.During the 1960s and the 1970s, Joel Moses demonstrated the power of symbolic reasoning for integration problems in the Macsyma program, a successful knowledge-based program first in the field of mathematics.In the 1990s marked a big acquisition in various fields of AI and demonstrations of various applications. More specifically Deep Blue, a chess computer, beat Garry Kasparov in a game 6 of the famous game in 1997. DARPA stated that the costs saved through the application of AI methods for scheduling unit in the first Gulf War had to replace the entire investment in AI research since 1950 on the U.S. government.Great DARPA challenge, which began in 2004 and continues to this day, is a race for $ 2 million prize where the vehicle driven without communication with humans, using GPS, computers and sophisticated sensor array, several hundred miles across the desert area is challenging.

The development of Artificial Intelligence
Jacques de Vaucanson Jacques Pierre Drotz
Starting around the 18th century as machines have become more complex, the effort has attempted to create human imitation. In 1736 a French inventor Jacques de Vaucanson (1709-1782) made a flute player engine size as a man who can play 12 tone melody. Not only this course, mechanics are able to move the lips and tongue actually to control the flow of air into the flute.In 1774 a French inventor, Pierre Jacques Drotz astonish the people of Europe with an automation is about a boy who can sit down and write a note book. This discovery was followed by others, the automation in the form of a sweet girl who can play the harpsichord. Overall it was a mechanical process that performs motion with predetermined.

Humans are still trying to create another machine. In 1769, the plains of Europe were surprised by a game of chess that can address the steps the chess game that has not been determined in advance. This machine is called Chess Automation and Maelzel made by Wolfgang von Kempelan (1734-1804) of Hungary. But the machine was finally burned in 1854 in Philadelphia United Serikat.banyak people do not believe in the ability of the machine. And a writer from the United States, Edgar Allan Poe (1809-1849) wrote a refutation of the machine, he and his friends were right, that the machine is a hoax, and in fact is not aoutomation, but it is a very good construction controlled by a reliable chess players who were hiding in it.Attempt to create a game engine construction continued in 1914, and the engine was first demonstrated chess engine. This machine is the inventor Leonardo Torres Quevedo Y, director of the Laboratorio de Automatica in Madrid, Spain. Several years later, the idea of ​​a chess game developed and implemented on a computer by Arthur L. Samuel from IBM and was further developed by Claude Shannon.In the 20th century, Automation has been developed and applied primarily in the United States armed forces, a battle simulation program. Today, the development of AI has reached the stage that can be said to be fantastic, especially in the following areas:- Game Playing- General Problem Solving- Natural Language Recognition- Speech Recognition- Visual Recognition- Robotics- And Expert Systems

Playing gamesPlaying games (game play) is a very popular field of AI in the form of a game between humans against machines that have the intellectual to think. Playing with computers is interesting, even to forget the main task is more important. Computers can react and respond to the actions provided by the opponent.
Playing games (game play) is the field of AI in the form of a game between humans against machines that have the intellectual to think. Playing with the computer are interesting even to forget the main task is more important. Computers can react and respond to the actions provided by the opponent. Thousands of kinds of computer games have been created and developed.Game is a computer game created by animation techniques and methods. AI game play is a very popular field in the form of a game between humans against machines that have the intellectual to think. Computers can react and respond to the actions provided by the opponent.One of the embedded computer AI for the game called Deep Blue. Deep Blue is a chess computer made by IBM first to win a chess game against a world champion (Garry Kasparov) in a chess tournament standards. His first victory (in the game or the first round) occurred on February 10, 1996, and is a very famous game.
Now the AI ​​has improved a lot the game more interesting, interactive, and the graphics are very nice. Coupled with advances in computer networking technologies that are getting faster, there are a lot of game-based online game AI. Not a few people are interested in the game today. They play a game to fill their time or their skills training in thinking.
Type of Game
 Detailed information = a game where the game in knowing all the steps that may occur from himself and from the opponent and the final outcome of the game. Example game: chess and tic tac toe

 Incomplete information: a game where the player does not know all the possible steps opponent. Examples of game: Poker Cards and Brigde because all the cards are not known by the players.
 History of Artificial Intelligence in GamesIn 1769, the plains of Europe were surprised by a game of chess that can address the steps the chess game that has not been determined in advance. This machine is called Chess Automation and Maelzel made by Wolfgang von Kempelan (1734-1804) of Hungary. But the machine was finally burned in 1854 in Philadelphia United Serikat.banyak people do not believe in the ability of the machine. And a writer from the United States, Edgar Allan Poe (1809-1849) wrote a refutation of the machine, he and his friends were right, that the machine is a hoax, and in fact is not aoutomation, but it is a very good construction controlled by a reliable chess players who were hiding in it.Attempt to create a game engine construction continued in 1914, and the engine was first demonstrated chess engine. This machine is the inventor Leonardo Torres Quevedo Y, director of the Laboratorio de Automatica in Madrid, Spain. Several years later, the idea of ​​a chess game developed and implemented on a computer by Arthur L. Samuel from IBM and was further developed by Claude Shannon.

Artificial Intelligence in GamesOne element that plays an important role in a game is the artificial intelligence. With artificial intelligence, elements of the game can behave like human beings as natural as possible.Game AI is an application to model the characters involved in the game either as opposed to, or supporting characters who are part of the game but did not come into play (NPC = Non playable Character). The role of artificial intelligence in terms of player interaction with the game is to use interactions that are natural human family is used to interact with fellow humans. Examples of media interaction is:• Vision (vision)• Voice (voice), speech (speech)• Movement of limbs (gesture)Artificial Intelligence for the formation of the game was also used algorithm, ie n-ary tree species to a structure. Implementation of the tree (tree) is usually called a game tree. Based on the game is a game tree structured artificial intelligence algorithms. Artificial Intellegence which is embedded in a game that make up the game tree analysis usually represents a condition or position of the game from the game as a node, and represent steps that may be performed as directed sides which connect these conditions to the child node (child) as a representation of a tree (tree) .However, such direct representation usually has a weakness, namely the representation of the data tree will be very wide and many. Possible for a computer system capable of performing calculations as much as any issue, but a big game and big tree gives some problems, such as the consumption of the memory process, substantial storage capacity and performance are less on a low spec game consoles. Because it was formed a few algorithms and simplification of a game tree.On one example of the classic game, that tic tac toe, simplification can be done by various methods. One of them is minimax. The method was successfully applied and give a significant reduction. And not only can be used as monotonous, minimax also be used for games such as chess is more complicated, of course with different algorithms and representations.Minimax, which is one method of implementation (implementation) n-ary tree in a game, indicating that the implementation of the structure (especially trees) is required in the manufacture and application of Artificial Intelligence, and did not rule out the science and new, more sophisticated methods will be found in the front.

Some of the characteristics and limitations of game for game playing:Played by 2 (two) players: humans and computers. The players take turns walking.A. Perfect Information GameBoth players have the same access to complete information about the state of the game, so no information is closed to the opponent.2. Determined by No ChancesNot involve the factor of probability, for example by using dice.3. No Phsychological FactorsDoes not involve psychological factors, such as "snapping" (eg Poker)4. No Oversight Errors. Smart OpponenOpponents also assumed to be smart, so do not expect the opposite mistake, resulting in one step.
Some examples of games that are commonly used as an example the case of Game Playintyle = "font-family: courier new;"> Last One Loses n• n-coins Grundy's Game• Slide-5• Tic-Tac-Toe• Checkers• Go• Nim• Othello• Chess
The rapid development of the game at this time also need something different to the game rule. A game system, if it has been played to completion by a player, then when the same player who started the game again from scratch, then the rule will be the same game. but different for games that already exist today. system in the game, can learn to recognize patterns of play and the player when the player starts the game again, then the system will use a different rule for these same players. so the game becomes more interesting and challenging to play.
Examples of applications of artificial intelligence in the form of the game very much at all, there is a form of PC games, and some are in the form of a network game. Examples of applications that game Tic Tac Toe game.
Tic tac toe game is a game that uses a board of size n rows and n columns so that the size of a nxn board eg 3 x 3.This game is a game that human thinking skills, in which each player must make the image vertically, horizontally, left lateral, and oblique right to obtain the value. If the player can not form the desired image formation is puz declared lost. And when the image patterns otherwise balanced the game or series drow. This game thinking skills so that the players must take into account what is good and the result of the action taken.

Using Heuristics in GamesThe gameplay is an important test-bed for heuristic algorithms. Two-person game is more complicated than a simple puzzle since they involve an unexpected opponent.Minimax ProcedureThe Game of Nim: A token is placed on a table between two opponents. At each of the players have to divide the pile into two piles token was empty of any size. So, 6 tokens can be divided into 5 and 1, 4 and 2, but not 3 and 3. The first player to move loses the game.For a small token search space can be searched thoroughly. The following figure provides a space for the full 7-game token.
 In a two-person game, you should assume that your opponent has the same knowledge that you are doing and apply as well as you do. So at every stage of the game you have to consider the other to make the best move available. This is the basis of the minimax procedure.In minimax, the players are referred to as MAX (players) and MIN (opponent). Both are trying to maximize their movement. MAX players, trying to maximize its value. And MIN is the opponent tries to minimize the score MAX.Minimax procedure on Complete Space SearchA. Label each level of the search space in accordance with that move it at that level.2. Starting at the leaf nodes, each leaf node labeled with a 1 or 0 depending on whether it is a win for MAX (1) or MIN (0).3. Climbing to the top: if the parent MAX, MAX gives his children.4. Climbing to the top: if the parent MIN, MIN gives his children.
Consider the graph minimax for the Nim game. The value in each state that represent the best country players can hope to achieve. The values ​​obtained are used to choose among alternative moves.
 Minimax heuristicFor most games are not likely to expand the graph to a leaf node. Instead the strategy n-move look-ahead is used. Extended state space to the level n. Each leaf node in this subgraph is assigned a value according to the heuristic evaluation function. Values ​​are then propagated back to the root node. Redistributable value represents the heuristic value of the best attainable state of the node.
 Example: Samuel chess program using a weighted sum as the evaluation function. It uses a simple learning algorithm to adjust the weights after wins and losses, so that the program improved over time.


Alpha-Beta procedureAlpha-beta pruning is a procedure to reduce the amount of computation and search for minimax. Minimax is the search for two-pass, one pass is used to assign values ​​to the node at a depth heuristic and the second ply is used to spread the values ​​of the tree.
Alpha-beta results in depth-first search. An alpha value is the value associated with the initial or temporary MAX node. Because the MAX node the maximum rated among their children, can not lower the alpha value, only go up. A beta value is the initial or temporary value associated with node MIN. Because the nodes are given the minimum value MIN of their children, the beta value can never increase, just get it down.For example, alpha node MAX = 6. Then search no need to consider each branch emanating from a descendant MIN has a beta value of less-than-or-equal to 6. So, if you know that have a MAX node alpha 6, and you know that one of the descendants of MIN which has a beta of less than or equal to 6, you need not look further below MIN node. This is called alpha pruning.The reason is that no matter what happens below the MIN node, can not take a value greater than 6. So its value can not be propagated to (alpha) MAX his parents.Similarly, if the MIN node's beta value = 6, you need not look further below the MAX descent who had obtained an alpha value of 6 or more. This is called beta pruning.The reason again is that no matter what happens below the MAX node, can not take values ​​less than 6. So its value can not be propagated to (beta) MIN his parents.Rules for Alpha-beta pruning• Alpha Pruning: Search can be stopped under any MIN node has a beta value less than or equal to the alpha value of each of its MAX ancestors.• beta Pruning: Search can be stopped at the bottom of every MAX vertex alpha value greater than or equal to the beta of its MIN ancestors eachh.



CONCLUSION

Artificial Intelligence (AI) or Artificial intelligence is created and put into a machine (computer) in order to do the job as do humans. Several kinds of fields that use artificial intelligence expert systems, among others, computer games (games), fuzzy logic, neural networks and robotics.
Broadly speaking, the AI ​​is divided into two schools of thought are: Conventional AI (AI symbolic) Computational Intelligence (AI non-symbolic)
CONCEPT OF ARTIFICIAL INTELLIGENCEA. Game (Game)2. NATURAL LANGUAGE3. Robotics and SENSOR SYSTEM4. EXPERT SYSTEM
Playing gamesGame is a computer game created by animation techniques and methods. AI game play is a very popular field in the form of a game between humans against machines that have the intellectual to think.Type of Game
 More information
 Incomplete information
Some examples of games:• Tic-Tac-Toe• Checkers• Othello• Chess• Card Poker







REFERENCES



T. Sutujo, Edy Suhartono Mulyanto and Vincent 2011 "Artificial Intelligence", New York: Publisher Brenda
 http://ai-a-erial.blogspot.com/2011/02/aplikasi-kecerdasan-buatan.html
 http://e-rara4mystudy.blogspot.com/2010/10/kecerdasan-buatan-game-playing.html
 http://hennytandiono.wordpress.com/2011/04/06/artificial-intelligence/
 http://id.wikipedia.org/wiki/Kecerdasan_buatan

 http://lovelyploeto.blogspot.com/2011/11/kecerdasan-buatan.html
 http://ndoware.com/kecerdasan-buatan.html
 http://setiyanugroho.wordpress.com/2011/04/12/kecerdasan-buatan-dalam-game/