END TERM EXAMINATION [MAY-JUNE 2015] SIXTH SEMESTER [BCA] ARTIFICIAL INTELLIGENCE [BCA-312] Time:3 hrs. Note: Attempt any five questions including Q. no. 1 which is compulsory. Select one question from each unit. Q. 1. Answer the following questions: (a) What is AI? Mention application of AI. (2.5) Ans. Refer Q. 3. (b) of End ‘Term Examination 2018. Q. L.(b) Name the two kinds of synchronous rules that allow deductions. (2.5) Ans. Rule- Based Deduction Systems- The way in which piece of knowledge is expressed by a human expert carries important information, example: if the person has fever and feels tummy-pain then she may have an infection. In logic it can be expressed as follows: Vx. (has_fever(x) & tummy_pain(x) > has_an_infection(x)) * Forward production system * Backward production system The main idea behind the forward/backward production systems is: 0 to take advantage of the implicational form in which production rules are stated by the expert © and use that information to help achieving the goal. In the present systems the formulas have two forms: o rules o and facts Natural Deduction- Testing whether a proposition is a tautology by testing every possible truth assignment is expensive—there are exponentially many. We need a deductive system, which will allow us to construct proofs of tautologies in a step by- step fashion. The system we will use is known as natural deduction. The system consists of a set of rules of inference for deriving consequences from premises, Q.1. (©) State the reasons why the inductive logic programming is popular: (2.5) Ans. Inductive Logic Programming (ILP)is a subfield of machine learning that learns computer programs from data, where the programs and data are logic programs. It may also be explained as a form of supervised machine learning which uses logic programming (primarily Prolog) as a uniform representation for background knowledge, examples, and induced theories. ILP is preferred over other machine learning approaches because of its easy comprehensibility, intelligibility, and ability to include additional information in the learning problem. In its earlier days, Inductive Logic Programming was applied to synthesizing functional or logic programs for general purpose tasks such as manipulating data structure (for example, sorting or reversing a list). These investigations showed that small programs could be synthesized from a few input/output examples. The recent IT revolution has created real-world opportunities for more such techniques and applications. Most computer users, today, are non-programmers and are limited to being passive consumers of the software that is made available to them. ILP can empower such users to more effectively leverage computers for automating their daily repetitive tasks. Inductive Logic Programming is widely used in the areas of End-user Programming and Education. Scanned with CamScanner2016, Sixth Semester, Artificial Intelligence Q. 1, () What ix constraint satinfactio Ang, In artificial intellig the pro ? neo and operations research, constraint satisfac 1s of finding 4 solution to Kel of constraints that impose the variables munt satinfy . A volution ia therefore a net of values for the variable (2.5) tion is conditions that a that sutinfion all constraints-that iw, a point. in the feanible re ‘The techniques used in conntraint satisfac boing considered, Often uned are constraints on a finite domain, to the point that constraint satinfuction problems arc typically identified with problems based on conatraints on a finite domain, Such problems are usually solved via search, in particular ‘m of backtracking or local search, Constraint, propagation are other methods used on much problem, most of them are incomplete in yeneral, that ig, they may solve the problem or prove it unsatiofiable, but not always, Constraint propagation methods are also uned in conjunction with search to ma problem simpler to solve. Other considered kinds of constraints are on real or rational numbers; solving problems on thene constraints in done via variable elimination or the simplex algorithm, Constraint satinfiaction originated in the field of artificial intelligence in the 1970s. During the 1980s and 1990u, embeddings of constraints into a programming language was developed cl patraint programming are Prolog and C++. QL (eo) Whatis tb intic Seareh? (2.5) Ans. A heuristic nethod that. + might not always find the best solution + But iv Lo find a good solution in reasonable time. * By nacrificing comploteness it increases efficiency. + Useful in solving tough problems which » Could not be solved any other way. -» Solutions take an infinite ion. jon depend on the kind of constraints ne or 'y long Lime to compute, ‘The classic example of heuristic search methods is the travelling salesman problem, Heuristic Search methods * Genorato and ‘Tost Algorithm + Hill climbing + Best First Seareh * Problem Reduction (AO* Algorithm) Q.1. (f) Defino Expert System Shell, Ans. Refer Q. 9. (b) of End ‘Term Examination 2017, Q. 1. Gp Discuss the advantages of Dee Ans, A decision support system (DSS) is business or organizational decision-making a operations and planning levels of an organization (2.5) ion support system, (2.5) information system that supports ics. DSSs serve the management, (usually mid and higher management) and help people make decisions ubout problems that may be rapidly changing and not easily specified in advance~i.c, unstructured and semi-structured decivion problems, Decisio? ‘support systems can bo either fully computerized or human-powered, or a combination of bot Tho advantages of decinion support system are: a8) * The clinical decision support system for medical diagnosis, There are four stages wolution of clinical decision support system (CDSS): the primitive version is standal"® and does not support integration; the second generation supports integration with ot? medical nystons; the third ix standard-baved, and the fourth is service model-based. * DSS is exter voly used ines He Ey is i DSS Is extensively used in business and imanagement, Exeeuti dashboard 8% other buinews performance software allow faster dees eee eee ratiol Hegative trond, and better allocation of business resoa co MANNE: identifies Begs Scanned with CamScannerLP, University-[BOA)-Alaal Books wh + A growing aren of DSS application, concepts, peiiedpiless, aad beebiiiq iin He in yricultural production, marketing for sustainable develypeieet DSS are algo provalent i forest miarnygement ere: Ue ton ghana and the spatial dimension of planning problems demands sperttic requlesnentn A aspects of Foi management, from log tranaportalion, hrerveat He hed ine bn austainability and ecosystem protection have heen adtlresaed by modern DEBE Q.1.(h) How does alpha bets i leohnique works? (Ahi) Ans. Alph ogee Hie rninbiar of nodes that are evalunted by the minimax algoritlin in ils pearel bree Hissin adder wnetieh search aljorithm used commonly for machine playing af Live-player gstnes (He bie too, Chess, Go, etc.). It stops evaluating a move whens Hes ate pssst bs he ni fond that proves the move to be worse than a previously Hrnisve, Heb meres need ti be evaluated further. When applied to a standard 1 He retuurnes Uh ssinies sree max would, but prunes away branches that etniol possibly infliestiie the fine denise The algorithm maintains tivo values, alpha and bets, ahieh representa the minimum score that the maximizing player is assured of wud (he inaximum seare Hat the minimizing player, is assured of respectively, Initially, slphi is negeeline intinity and beta is positive infinity, i.v. both players stark with their worst possible suare Whenever the maximum score that the minimizing player (ie. Whe “bebo” player) is assured of becomes less than the minimum score that Lhe minnizing player (ie, Wie “alpha” player) is assured of (i.e. beta s alpha), the maximizing player need nol consider further descendants of this node, as they will never be rertched in the aetual plivy, oy 1 is a search algorithien Uyal. seeks bo te bon renee Above figure shows the alpha-beta pruning. The grayed-out sub trees need not be explored (when moves are evaluated from Jeft to right), vines we know the group of sub trees as a whole yields the value of an equivalent sub Lew or worse, and 9s auch cannot, influence the final result. The max and min levels represent the turn of the playar and the adversary, respectively. Q.1.() What is Local minima problem? (2.5) Ans. A local minimum of a function (typically a cost function in sachine learning, which is something we want to minisize based on empirical data) is a point in tha domain ofa function that has the following property: the function evaluates tou greater value at every other point in a neighbourhood around the locul minimnusn (a neighbourhood in this case can correspond to a “ball” around the minimum) than the local minimum itwelf, ‘The graph of an elementary (1D) function: : ~~ G\ot0! maximum toch nasean 7 Jaca mines ~~ gicval sninntium bhbonae T 04 08 08 4 42 Scanned with CamScanner| Ninth Semester, Artitivial Sntetligenee __Bandhlocal minima in an array: Chien wn werey wt hf Miatine incagern, the task is to tind a local minimu in it. We say Uhl an element asx, me a lenak moveneervena igh is fess than ae equal ta both its neighbors + For carner olements, we need tn vonsider uly one niin tor osmopretann, + -‘Thereeanhe more than one loc) minisns in en were, ms tea Sets many es Ct Examplesi Anput: ari = 19, 6,8, 145,145 Output: Index of local minivnss is 2 "The output prints index of % herause it ix aencllier thier inthe oh sta eecgttirwers Note that indexes of elements 6 and 4 ure alse valid vig Q. 1G) Distinguish between Predicate and Propesiticnal logic. (25) Ans. In propositional logic, a stutenens that can either lve trun or Lala ie caled proposition. For example, the st it's reinitay watt” in eithaes trae, 0 Cali ‘This statement would be translated inte progeminsunsl Lapel inesuaage es = cial attoet like P. Ifyou have one or snore pens nine teat tik sentences using logien! wunnerti vies Iie "7 whi west 7 In symbols these ounnectives link lie this + natn + ander * ony then: = + fend unly Bex In predicate logic, ye have everything that exsete in propasitianal logic, test nam you have the cbiliey vs eturibute properties and relationabspe on things or variatdes. A L place predicate i 2 statement that says wmething dhoat an shipet An exasnphe fh this would te “two is an even number ” This etatenens bs esxyinig tne tuentoar tore has thee ‘We can alts wae vatiahles that range over chyents, twat arent , nage tbactnsschvce. As wate teas weve be" i es ever rater.” | Now the fret eneremes was true alvin ting, but the sectad satan. is ony tewe if % stands in for in even number. In the predicate languays we cat represent these ae: * Pine xisen even nuvenber + Let acts, then Yan two is an even musnber ‘These I place gece cast ooncweted with tne ications trae peng inion gfe. Now, you wanted to exy that everything is bbe. In indicate the idea f everything we use the eymabol A stateraent thet tases this eymted ts walled a wn quastibiel “ : arse waly | Satexaent. Also, whhss if yo wate Uo wy that ners ute ja sonne tious that are Hise le usually sey, “here exists is thing thst isthe" The eyribed we was: tor there exiat ie 3.4 ssetement with thot eaind js called en eaetantally quattitind watararts hen we wee Uses in a logies sentence, We pr the variable to tbe right. of these ae then 3 preticate with that variabhe mex ta it, ike: . x * Bx ° Finelly ween sey A propasition is a statenaent that in either trae or false, A predicate 08 2 proyssition thet contains une or murs variables oe paramere ie ines warts, 2 predivate isn poramueerines prousttion, P:2 is en ineeger Prsisste, Px): xis 9n imager Scanned with CamScannerLP. University-[BCA]-Akash Books 2015-5 Q.2. (a) “Al is both a science and a field of engineering”. Discuss wich the help of an example. S 6. Ans. Artificial Intelligence is both a science and engineering. Artificial Intelligence attempts to mimic the working of the human brain. So, it combines the science of how the human brain works and the engineering behind machines. : Artificial Intelligence is both science and engineering of making intelligent computer Programs, it has been demonstrated that computers can be programmed to carry out, very complex tasks—as, for example, discovering proofs for mathematical theorems or playing chess—with great proficiency, Artificial intelligence (AI) is an area of computer science that emphasizes the creation of intelligent machines that work and react like humans, activities like: * Speech recognition * Learning + Planning _* Problem solving All activity will be done the way human does so we can say in AI we use science and engineering both. Q. 2. (b) Explain Tic-Tac-toe problem. (6.25) Ans. Tic-tac-toe is a pencil-and-paper game for two players, X (ascii value 88) and O (ascii value 79), who take turns marking the spaces in a 3x3 grid. The player who succeeds in placing three respective marks in a horizontal, vertical, or diagonal row wins the game, Empty space is represented by(ascii value 95), and the X player goes first. Here is an example game won by the first player, X: AEE SHE SEE SHE SHE SHB Sus ‘The function nextMove takes ina char player, the function to print2 space separated integers r and c which denote the row and calving that will be marked in your next move. The top left position is denoted by (0, 0). Q. 3. Explain the following search algorithms: (a) Depth first search Ans. Refer Q. 1. (g) of End Term Examination 2018, Q.3.(b) AO* Algorithm Ans. Step 1: Place the start node s on open Step 2: Using the search tree constructed thus far, solution Tree T,, Step 3: Select a node n that is both on open and a part of Ty, Remove n from open and place it on closed. : Step 4: Ifn is a terminal goal node, label n as solved. Step 5: Ifn is not a solvable node, label n as unsolvable. If the start node is labelled as unsolvable, exit with failure. and the 3x3 board as an array. Complete (6.25) (6.25) compute the most promising Step 6: Otherwise, expand node n generating all of its successors. For each such togive indienh 2a that contains more than one subproblem, generate their successors ite preden idual subproblems, Attach to each newly generated node n hack pointer to NS predecessor, Compute the cost estimate h* for exch newly generated node. Step 7: Return to step 2. Itcan be shown that AO* the efficiency depends on how Qa will always find a minimum-cost solution tree. Like A‘, closely h* approximates h. (a) Illustrates the use of First-order-logic to represent the knowledge, 5) Scanned with CamScanner6-2015, Sixth Semester, Artificial Intelligence _ First-order logic is symbolized reasoning in which each sentence, oF statement, broken Son, intoa: subject onda predicate. ‘The predicate modifies or a the properties of the subject. In first-order logic, a predicate can only refer to a single su ject First Togic is also known as first-order predicate calculus or first-order functional o& etc ‘A centeneo in first-order logic is written in the form Px or P(x), where Pis the pre : ae and xis the subject, represented as a variable. Complete sentences are logically combine and manipulated according to the same rules as those used in Boolean algebra. In first-order logic, a sentence can be structured using the universal quantifier (symbolized V) or the existential quantifier (3). Consider a subject that is a variable represented by x. First Order Logie « ‘The prepositional logic only deals, with the facts that may be true or false. — «+ ‘The first order logic assumes that the world contains objects, relations and functions. Syntax for first order logic: In prepositional logic, every expression is a sentence that represents a fact. First order logicincludes the sentences along with terms which can represent the objects. + Constant symbols, variables and function symbols are used to build terms, while quantifiers and predicate symbols are used to build the sentences. Syntax: Constants ae Functions Size, Color Variable xa ‘Terms Constant, variable or function(Term..) Predicates ‘True, False Quantifiers v3 Atomic sentences Predicate, Predicate(Term,..), Term=Term Sentences + Sentence, Senterite v Sentence, Sentence Sentence, : Sentence <> Sentence, Sentence «> Sentence, Quantifier Variable,...Sentence Semantics: Lets understand with an example: Consider the sentence “Blephants are big”. ‘There are many ways to represent this sentence. HasSize(Blephant, Big) SizeOF(Elephant)= Big Lets introduce a new syntax: IsEqual (SizeOf(El ig), thi object Elephant is big, which is a useless fact in aad cba eecear ator ake Elephants in general. So let’s represent that all Elephants are big So, we can find FOL statement as, " * All things that are Elephants are big. For all things x, for which x is a Elephant, x is bij ° For all things x, ifx is a Elephant, then xis bi ms Finally the FOL will be written as, | Vx Elephant (x) e> Big(x) Knowledge Engineering of FOL e knowledge engi oi and relationshiae mut KNOW about the domain to represent the important The following ste} . 1. Identity thep as ane ea to develop the knowledge base: eo. Scanned with CamScanner1.P. University-[BCA]-Akash Books 2015-7 2. Assemble the rel evant knowledge to the given problem or task. 3, Decide on a vocabulary of predicates, functions and constants. 4, Encode the general knowledge about the domain and a description of the specific problem instan 5. Apply queries to the inference procedure and get the answers. 6. Finally, debug the knowledge base. © Q44. (b) Describe mapping between facts and repr entation. iz Ans, {n all variety of knowledge representations, we deal with two kinds of entities, "Truths in some relevant world, These are the things we want to represent. B. Representations of facts in some chosen formalism. These are things we will actually be able to manipulate. ¥y lo think of structuring knowledge level these entities is at two levels: , at which facts are described, and 'ymbol level, at- which representations of objects at the knowledge level are defined in terms of symbols t} manipulated by programs. ‘The ntations are linked with two-way mappings. This link is ealled tation mappings. The forward representation mapping maps from facts to Ferreeentations, ‘The backward representation mapping goes the other way, from Tepresentations to facts, One common repr Regardless of the represe: concerned with an Engli information into and o senteng to the represi Q.5. (a) knowledge? (7.5) Ans. Inheritable knowledge: Relational Knowledge is made up of objects consist * attributes * Correspondin, We ex (particularly English) sentences, Jjration for facts we use in a program, we may also need ee be sh representation of those facts in order to feeilitars getting ut of the system, We need mapping functions from English entation we actually use and from it back to sentences, Differentiates between Inheritab! 0 if associated values, nd the base more by * Property inheritance * Elements inherit values * Data must be organised into a hierarchy of classes (Fig. 1) y allowing inference mechanisms: age A Isa ™ (Avant Gardel Jazz instance instance ‘ Miles Davis John Zom bands bands Miles Davis Group Naked City [Miles Davis Quintet Massada ig 1. Property Inheritance Hierarchy Scanned with CamScanner6-2015 Sixth Semester, Artifi voningin whieh ach sentence, or statement i hiteate, The predicate modifies or defines the properties broken down into subject te ats eum only refer tou single subject. First-order of the subject, In first ord ter predicate calculus or first-order functional calculus. logic is also known "ordor logic ia written in the form Px or P(x), where Ps the predicate Asontenty ject represented as a variable. Complete sentences are logically combined ad sipulated according, to the same rules as those used in Boolean algebra, In firstorder logic, a sentence can be structured using the universal quantifier (aymbolized V) or the existential quantifier (3), Consider a subject that is a variable represented by X, First Order Logic + The prepositional logic only deat + The first order logic assumes that the world contain Syntax for first order logie: In pre that represents a fact + First orderlogicincludes the sentences along with terms which can represent the objects. * Constant symbols, variables and function symbols are used to build terms, while quantifiers and predicate symbols are used to build the sentences. Ans. First-order logic is symbolized .with the facts that may be true or false. s objects, relations and functions. every expression is a sentence ‘Syntax: Constants A,B,C... Functions Size, Color Variable xa Terms Constant, variable or function(Term..) Predicates ‘True, False Quantifiers v3 Atomic sentences Predicate, Predicate(‘Term,..), Term=Term Sentences > Sentence, Sentente'v Sentence, Sentence a Sentence, Sentence <> Sentence, Sentence ¢> Sentence, Quantifier Variable,...Sentence Semantics: Lets understand with an example: Consider the sentence “Elephants are big”. There are many ways to represent this sentence. HasSize(Elephant, Big) SizeOF(Elephant)= Big Lets introduce a new syntax: IsEqual (SizeOf (Elephant, Big), this states that a object Elephant is big, which is a useless fact in any reasoning process about the Elephants in general. So let's represent that all Elephants are bi So, we can find FOL statement as, * All things that are Elephants are big. * For all things x, for which x is a Elephant, x is big. * Forall things x, ifx is a Elephant, then x is big. Finally the FOL will be written as, Y x Elephant (x) ¢> Big(x) Knowledge Enginccring of FOL The knowledge engineer must kno akjeots and relationships. ‘The following steps ace used to develop the knowledge base: F. Identify the problem or task, ov. 'w about the domain to represent the important Scanned with CamScannerLP. University-[BCALAkash Books 2015-7 2, Awwomble the rel 3. Decide on a vocabulary of predic 4, Encode the general knowled problem inetanes. 5. Apply queries to the inference procedure and get the answers. 6, Finally, debug, the knowledye base, Q-4, (hb) Describe mapping between facts and representation, (5) Ann, In all variety of knowledge reprenentations, we deal with two kinds of entities. A. Foetn: Truths in some relevant world, These are the thinigs we want to represent, J, Representations of facts in sone chosen formalism, These are things we will actually be able to manipulate, One way Lo think of structuriny (2) the knowledge level, ed, and (b) the xymbol level, at, which representations of objects at the knowledge level are ed in terme of iymbols hat can be manipulated by programs, els and reps inked with two-way mappings. This link is called 1 inappings. The forward representation mapping maps from facts to tions. The backward representation mapping goes the other way, from ions to facts. nt know qe: to the given problem or tank , functions and constants, bout the domain and ption of the specific entities is at two levels: representa roprese represen {ation is natural language (particularly English) sentences. tation for facts we use in a program, we may also need to be 1 English representation of those facts in order to facilitate getting information into and out of the system, We need mapping functions from English sentences to the representation we actually use and from it back to sentences. ntiates between Inheritable knowledge and Inferential (7.5) made up of objects consisting of Ans, Inheritable knowledge: Relational knowledge * attributes * Corresponding associated values. We extend the base more by allowing inference mechanisins: + Property inheritance + Elements inherit values from being members of a class. * Data must be organised into a hierarchy of classes (Fig. 1) age don Wake e_] iva Garde a Jazz ¥ “frase inetonco : [ tiles Davis John Zom ] bands bonds Scanned with CamScannerSixth Semester, Artificial Intelligence * Boxed nodes — objects and values of attributes of objects. ‘Values can be objects with attributes and so on. Arrows — point from object to its value. ‘This structure is known as a slot and filler structure, semantic network or a 8-2015 collection of frames. ‘The algorithm to retrieve a value for an attribute of an instance object: 1. Find the object in the knowledge base 2, If there is a value for the attribute report it 3. Otherwise look for a value of instance if none fail 4, Otherwise go to that node and find a value for the attribute and then report it 5. Otherwise search through using isa until a value is found for the attribute, Inferential Knowledge Represent knowledge as formal logic: ‘ All dogs have tails 3: dog(x) > hasatail(x) Advantages: * Aset of strict rules. > Can be used to derive more facts. > Truths of new statements can be verified. — Guaranteed correctness. * Many inference procedures available to in implement standard rules of logic. + Popular in AI systems. Z.g. Automated theorem proving. Q.5. (b) Discusses Inheritable knowledge by applying the steps to real world application of your choice. oy ‘Ans, Inheritable knowledge: Relational knowledge is made up of objgets consisting of * attributes * Corresponding associated values. ° ‘We extend the base more by allowing inference mechanisms: * Property inheritance > Elements inherit values from being members of a class. -» data must be organised into a hierarchy of classes. Inheritable knowledge: * Objects are organized into classes and classe’ are organized in a generalization hierarchy. ‘= Inheritance is a powerful form of inference, but not adequate. + Ex. Property inheritance inference mechanism. handed Aduttmale -—S3__4} person Right instance Peter >__ Tata Consultar Works_at = @.6.(a) Explain syntactic processing, (6.28) Ans. The most basic buildin, e ig blocks of langua, li bee large, but Brite, set of words, Words are formed ints pied Reeetence then, Ute jrellformed sequence of words. The language is the set of all sentences that are well” formed, i.e., that follow a set of rules. This set of rules is called a grammar. Scanned with CamScanner1.0. University (ICAL Akash Books 2015-9 ‘Wy do ayitaetio analysis, we build. a parser, ie. a sollware systems that checks Whether (he ruler ate followed and that phivides atv analysis based on the grammar VSeutenve ss NP VP Gie.ca sentence ean be firmed ftom a noun phrase followed by avery pstieame, . PNY > HOI (A HOME phage can be jist a single noun) SND os oC non (a noun phrase eat be a determiner followed by a noun) ANT 5 yort (a verb phrase ean be just a single verb) ayy > verb NV (a vert phrawe ean bo a verb followed by enivnow ix The ont ate the fish . We calla parse (ree: # noun phrase) w let oun verb S: the al ate det Hex ‘Dhis might be represented textually as: Wontonce une (alot the) ‘(noun eat), we . (verb ate) une (dot tho) (noun fieh)) Q.6. 00) Differentiate between xyntactic process and semantic processing. (6.25) Ana, Rotor Q. 7. (0) of Bnd Torm Examination 2018, Q. 7 Write short note on the following (0 Explanation Based Learning Ans, Exphination-Based Learning (EBL) isu form of machine learning that exploits a very xlrong, or even perfect, domain theory in order to make generalizations or form concopts thom training examples, Good application domain for an EBL is natural Janguage processing (NLP), An example of BBL using a perfect domain theory is a program that learns to play choas through examplo, A specific chess position (hat contains an important feature auch ad “Vorcod loss of black queen in two moves” includes many irrelevant features, tuch au tho specific seattoring of pawns on the board, EBL can take a single training: oxnmplo nnd dotormine what the relevant foutures are in order to form a generalization, TEL nyatom works by finding a way to deduce each training example from the gysten's extating database of domain theory, Having a short proof of the training example extends the domuin-beory database, onabling tho EBL system (o find and classify future examples that uno sitnilar to tho training example vory quickly. the main drawback of the method—the cost of applyinys Che learned prot macros, as those become numerous-—-was analyzed by Minton, EBL software takes four inputs: + A hypothouls apace (the sot of all possible conclusions) + Adomuain theory (axioma about a domain of interest) (6.25) Scanned with CamScannerester, Artificial Intelligence 10-2015, Sixth Set © ‘Tyaining examples (specific facts | ow + Oporntionality criteria (criteria for determining are efficiently recognizable, et. which fealuras are re Qu1.(b) Learning by analog, . We all learn new knowledge: ned, the amount of rel facts that rule out some possible hypothesis) + which features in Uhe domain Uy detectable using sensors) (6.25) depending on the type of evant knowledge we already possess, and the ning takes place. There are five methods of learn They are, ‘ote learning) 2, Direct instruction (by being told) 4, Induction through different methods, Deduction The third type lis solution through the t when solving problen Jojgical learning, is the process of learning a new concept or ofsimilur known concepts or solutions. We use this type of learning 1 examination where previously learned examples serve as a guide or when we learn to drive a truck using our knowledge of car driving. We make frequent uso of anal 1 of learning requires still more inferring than either of the previous form iL transformations must be made between the known and unknown situations. This is a kind of application of knowledge in a new situation. Q.8. Write short no(es on the following (any two): (6.25 x 2 = 12.5) (a) Expert system (b) Concept of uni nly in Expert system (c) LISP Programming Ans, (a) Expert system- Rofer Q. 9. of End Term Examination 2018. (b) Concept of uncertainty in Expert system im of artificial intelligence is to develop tools for representing piece of knowledge and providing inference mechanism for elaborating conclusion of knowledge from stored 'The available knowledge is far from being certain, precise and complete. 1 systems the word uncertainty is related to the working with inexact data, imprecise information, handling identical situation, reliability of the results ete. An expert system allows the user to assign probabilities, certainty factors, or confidence leva ae techniques eae all input data. This feature closely represents Bis feet bre bleia oe ene a id. An expert system can take all relevant a a 4 recommendation based on the best possible solution rather than the only exact solution to handle such problems. ‘TYPES OF UNCERTAINTY i faite Por erate i. iu Lert to determine whether an assertion in the model is true aise ae mn ighth uncertainty about the fact ‘the height of plant is 38”. in fermumpla tere cuns akan ete an ne rul onan apes shad Hhtof plant is betwe q' , the information available might be her 37 or 43°), negative (e.g., (often referred to as incompletene a to) Vi are inher formal n37 and 43”), disjunctive (eg, “the height of height of plant is not 37”), or even unknown mode inetd elements (eg, predicates or quantifiers) that , a is in i ‘ i he model contain {wo or more agsertions that cannot be true at the ae — 37 and 43” and “the height of plant is 35.” ¢ model lack complete semantics, leading t Tpretations. For ex: i ation. For ‘ample, it may not be clear whether stated (©) LISP Progcamming. tion of vaguer (@) Inconsistency same time: (e) Aimbiguity—Some ¢ several possibl temperature is in Refe 7 ‘afer Q. 8, of End Term Examination 2018. Scanned with CamScanner