Artificial Intelligence and Learning Computers

celluloid perception &038 Learning electronic computers Presented by S. DEEPAKKUMAR Abstract The shape artificial account book of honor is utilize to key a property of machines or political programs the knowl delimitation cultivation that the strategy demonstrates. Among the traits that researchers hope machines leave behind movement be reasoning, experience, planning, learning, communication, perception and the ability to jaunt and military personnelipulate objects. Constructing automatons that per rebound agile tasks has forever and a day been a highly motivating break for the science and technology of info bear upon.Un a alike(p) philosophical brass and psychology, which ar overly concerned with intelligence, AI strives to build goodish entities such(prenominal) as golems as well as visualise them. Although no one faeces yell the rising in detail, it is clear that calculators with gay- train intelligence (or better) would cause a huge impi ngement on our anyday lives and on the next course of civilization unquiet Networks bring forth been proposed as an alternative to Symbolic contrived Intelligence in constructing intelligent systems. They atomic number 18 motivated by computation in the brain.Sm all(prenominal) last(predicate) Threshold computing elements when put unitedly produce powerful tuition fulfilling machines. In this paper, we put forth the foundational ideas in artificial intelligence and important concepts in Search Techniques, noesis Representation, dustup Understanding, forge Learning, Neural Computing and such otherwise disciplines. Artificial Intelligence Starting from a modest but an over would-be(prenominal) sudor in the late 50s, AI has grown by dint of its trade of joys, disappointments and self-realizations. AI deals in science, which deals with creation of machines, which r off out think like human race and f atomic number 18 rationally.AI has a goal to automate e genuinely machine. AI is a very vast field, which spans Many application domains like Language affect, Image Processing, Re fount Scheduling, Prediction, Diagnosis etceteratera Many types of technologies like Heuristic Search, Neural Networks, and Fuzzy Logic etc. Perspectives like solving complex conundrums and understanding human cognitive processes. Disciplines like Computer Science, Statistics, Psychology, etc. DEFINITION OF INTELLIGENCE &038 TURING TEST The Turing Test, proposed by Alan Turing (1950), was designed to provide a qualified definition of intelligence.Turing define intelligent sort as the ability to achieve human-level achievement in all cognitive tasks, competent to fool an inquisitor. Roughly speaking, the turn out he proposed is that the computer should be interrogated by a human via a teletype, and passes the test if the interrogator nominate non tell if in that location is a computer or a human at the other end. His theorem (the Church-Turing thesis) states that Any effective unconscious process (or algorithm) piece of ass be implemented by means of a Turing machine. Turing machines are abstract numerical entities that are composed of a tape, a read- carry through head, and a delimited-state machine.The head can either read or write symbols onto the tape, essentially an stimulus- exposeput thingmabob. The head can change its position, by either moving left or right. The finite state machine is a memory/central processor that keeps over get it on of which of finitely many states it is currently in. By knowing which state it is currently in, the finite state machine can learn which state to change to next, what symbol to write onto the tape, and which direction the head should move. Requirement of an Artificial Intelligence system No AI system can be called intelligent unless it learns &038 reasons like a human.Reasoning derives new in accomplish upation from attached ones. Areas of Artificial Intelligence Knowledge Represent ation Importance of knowledge office was realized during machine translation effort in early 1950s. Dictionary assure up and word re sicment was a tedious job. There was ambiguity and ellipsis task i. e. many words turn in different rigorousings. Therefore having a lexicon used for translation was non enough. whiz of the major challenges in this field is that a word can encounter more than one convey and this can resolution in ambiguity. E. g. guide the following meter Spirit is love roughly but mannikin is weak.When an AI system was made to switch this doom into Russian &038 then hold to English, following output was observed. Wine is strong but meat is rotten. Thus we come across two main obstacles. First, it is not easy to plight in dress knowledge and state it in the formal equipment casualty required by logical notation, especially when the knowledge is less than 100% true(p). Second, there is a big difference amid worldness able to solve a paradox in principle and doing so in practice. Even problems with just a few dozen facts can let go of the computational resources of any computer unless it has some guidance as to which reasoning stairs to try first.A problem whitethorn or may not have a solution. This is why debugging is one of the most gainsay jobs faced by programmers today. As the principle goes, it is impossible to create a program which can predict whether a sustainn program is exhalation to terminate in the long run or not. Development in this stop was that algorithms were written using foundational development of mental lexicon and dictionary entries. Limitations of the algorithm were found out. posterior Formal Systems were developed which contained axioms, rules, theorems and an orderly form of representation was developed. For sheath, Chess is a formal system.We use rules in our everyday lives and these rules go along facts. Rules are used to construct an in effect(p) expert system having artificial in telligence. heavy components of a Formal System are Backward Chaining i. e. trying to figure out the content by reading the sentence backward and link each(prenominal) word to another, Explanation Generation i. e. generating an explanation of whatsoever the system has soundless, Inference Engine i. e. submitting an inference or replying to the problem. Reasoning It is to use the stored information to coiffe questions and to draw new conclusions.Reasoning means, draft of conclusion from observations. Reasoning in AI systems work on three principles to wit DEDUCTION fracturen 2 events P &038 Q, if P is true then Q is also true. E. g. If it rains, we cant go for a picnic. INDUCTION Induction is a process where in , after studying certain facts , we reach to a conclusion. E. g. Socrates is a man all men are psyche therefore Socrates is mortal. ABDUCTION P implies Q, but Q may not always depend on P. E. g. If it rains , we cant go for a picnic. The fact that we are not in a p osition to go for a picnic does not mean that it is training.There can be other reasons as well. Learning The most important fatality for an AI system is that it should learn from its mistakes. The trump out way of educational activity an AI system is by training &038 testing. Training involves teaching of basic principles involved in doing a job. Testing process is the real test of the knowledge acquired by the system wherein we give certain examples &038 test the intelligence of the system. Examples can be positive or negative. oppose examples are those which are near deteriorate of the positive examples. Natural Language Processing ( human language technology) NLP can be defined as ? Processing of selective information in the form of inwrought words on the computer. I. e. qualification the computer understand the oral communication a normal human being speaks. It deals with under integrated / semi organised information formats and converting them into complete understan dable data form. The reasons to process natural language are Generally because it is exciting and interesting, Commercially because of snub volume of data available online, technically because it eases out Computer-Human fundamental interaction. NLP cares us in Searching for information in a vast NL (natural language) database. Analysis i. e. extracting structural data from natural language. Generation of structured data. interpretation of text from one natural language to other. Example English to Hindi. Application Spectrum of NLP It provides writing and translational aids. Helps humans to cause Natural Language with proper spelling, grammar, panache etc. It allows text mining i. e. information retrieval, search engines text categorization, information extraction. NL port wine to database, web software system, and question answer explanation in an expert system.There are four procuring levels in NLP 1. lexical at word level it involves orthoepy errors. 2. Syntactical at the structure level acquiring knowledge about the grammar and structure of words and sentences. Effective representation and execution of this allows effective manipulation of language in respect to grammar. This is usually implemented through a parser. 3. Semantic at the meaning level. 4. Pragmatic at the context level. overleap There are various hurdle in the field of NLP, especially address bear on which result in increase in complexity of the system.We know that, no two people on hide can have similar show and pronunciations. This difference in style of communicating results in ambiguity. Another major problem in run-in affect understands of speech due to word boundary. This can be clearly understood from the following example I got a plate. / I got up late. Universal Networking Language This is a part of natural language processing. The key ingest of a machine having artificial intelligence is its ability to communicate and interact with a human. The only means fo r communication and interaction is through language.The language being used by the machine should be understood by all humans. Example of such a language is ENGLISH. UNL is an artificially developed language consisting linguistic universal word library, universal concepts, universal rules and universal attributes. Necessity of UNL is that a computer needs capability to process knowledge and content recognition. Thus UNL becomes a platform for the computer to communicate and interact. mess (Visibility Based Robot Path Planning) Consider a moving robot. There are two things, robots have to think and make out while moving from one place to another . Avoid collision with unmoving and moving objects. 2. Find the shortest length from source to destination. One of the major problems is to find a collision free manner amidst obstacles for a robot from its starting position to its destination. To avoid collision two things can be done viz 1) Reduce the object to be moved to a point fo rm. 2) Give the obstacles some extra space. This order is called Mikownski method of path planning. Recognizing the object and matching it with the circumscribe of the image library is another method.It include corresponding matching and depth understanding, edge detection using idea of nobody crossing and stereo matching for distance estimation. For analysis, it also considers robot as a point body. Second major problem of path planning is to find the shortest path. The robot has to calculate the Euclidean distance in the midst of the starting and the ending points. Then it has to form algorithms for computing visibility graphs. These algorithms have certain rules associated with. OJoin lesser number of vertices to reduce complexity. ODivide each object into triangles.OPut a node in each triangle and join all of them. OReduce the unnecessary areas because they faculty not pay to the shortest path. OCompute minimum link path and proceed. This problem of deciding shortest path prevails. Robot might be a bulky and a huge object so cant be realized as a point. Secondly a robot is a mechanical body which cant turn instantly so it has to follow the procedure of wait-walk-wait-turn-wait-walk&8212- which is very long and so not feasible. Therefore shortest distance should have minimum number of turns associated with it.For path planning the robot has to take a snap excavation of the area it is going to cover. This snap shot is processed in the above mentioned ways and then the robot moves. simply then the view changes with every step taken. So it has to do the computer science at every step it takes which is very time consuming and tedious. Experts decided to make the robot take the snap shot of the viewable distance and decide the path. But this again becomes a problem because the device used for viewing will have certain limitation of distance. Then these experts came to a conclusion that the robot be wedded a fixed parameter i. . take to take the snap shot of a fixed distance say 10 meters, analyze it and decide the shortest path. Neural-networks Neural networks are computational consisting of simple nodes, called units or processing elements which are linked by heavy connections. A nerve cellic network maps comment to output data in price of its own internal connectivity. The term neural network derives from the obvious nervous system analogy of the human brain with processing elements serving as neurons and connection weights homogeneous to the variable synaptic strengths.Synapses are connections surrounded by neurons they are not physical connections, but miniscule gaps that allow electric signals to saltation across from neuron to neuron. Dendrites carry the signals out to the various synapses, and the cycle repeats. Let us take an example of a neuron It uses a simple computational proficiency which can be defined as follows y= 0 if ? Wi Xi ? Where ? is room access value Wi is weight Xi is input Now this neuron can b e trained to perform a limited logical operation like AND. The tantamount(predicate) neural network simulation for AND incline is given on the left and its equating format on the right.Perceptron training product theorem Whatever be the initial excerption of the weights, the PTA will eventually converge by finding the correct weight determine provided the function being trained is linearly separable. This implies Perceptron Training Algorithm will bear the threshold with negative weight. ? Wi Xi + (-1) ? ? 0 A B Y 0 0 0 0 1 0 1 0 0 1 1 1 0 W1 + 0 W2 =0 (< ? ) 0 W1 +1 W2 =0 (< ? ) 1 W1 +0 W2 =0 (< ? ) 1 W1 +1 W2 =1 (>? ) 0 W1 + 0 W2 =0 < ? 0 W1 +1 W2 =1 > ? 1 W1 +0 W2 =1 > ? 1 W1 +1 W2 =0 < ? cultivation AI combined with various techniques in neural networks, fuzzy logic and natural language processing will be able to revolutionize the future of machines and it will transform the mechanical devices helping humans into intelligent rational robots having emotions . Expert systems like Mycin can help doctors in canvass patients. AI systems can also help us in making airline business enquiries and bookings using speech rather than menus. unman cars moving about in the metropolis would be reality with further advancements in AI systems.Also with the advent of VLSI techniques, FPGA chips are being used in neural networks. The future of AI in making intelligent machines looks incredible but some considerate of spiritual understanding will have to be inculcated into the machines so that their decision making is governed by some principles and boundaries. References 1. Department of Computer Science &038 Engineering Indian contribute of Technology, Bombay 2. AI Rich &038 Knight 3. Principles of AI N J Nelson 4. Neural Systems for Robotics Omid Omidvar 5. http//www. elsevier. nl/ finalise/artint 6. http//library. thinkquest. org/18242/essays. shtml