i Question Bonk cat) i p \(Define Knowledge. reprarentation in the: coment: of AMT ond 'expiain TRE Ng ni con: a} in At gystero ‘(choi Now ledge Representohan— Knowledge representation ia AL +o | thar involves encading in Fo tmatHor abou ine eoorld in a Form that a cempusen {pase can utilize +o Solve Complex tagles cuc ag diagnosing a problerd understanding Natura langdage of planning actions. Ee — Knot ledge. reprementono reeang think of it ok a oy for At eysiems tov! Know! Phings Gnd ‘moke clecisions baged on that leclae, : : —i kno i tS _lae sign ‘cance fA AL— 7 H = 195 Gaderstonding infee maton] Facts 3 Knecledge teprebentosiodn hetpe AU “Systeme +5 Undewtond betud) factS about the ‘word such og relationship een Ai Rerent concepts : t At 2) Making ‘deciSiong— BY representing Knowledge, tr) anolyre Wa fooler nC odion i Hake decicans For €R, a Gmowt assictant con nelp yeu find Food o restaurant by Knowing wnat + pes oF | Jou ike Vand Your lecaton. 3) Ledisning 5 Ad Gyolerm con learo. from new informotea and dote sheie Knowledqe: | Dose). for “example, VP a CeLf-driving Sam _j encoliniens a. ineud TLRS of obstacle THecao a {eon Rew te deal pith ib bowed on i418 — ND: pledge, ae Saacct eee = 0 hs Problen éo knowledge pam QMows. At. wing Raesot Solve “camphor a PL ise ex. IN a game [ike eh ae ee —~ [needs to vepWrent OU poss: Sci Ne ook Lion. — cutcemes tis make ihe best deat Sa De eine on “Fae gons _noere “eR Rech yelp eX ~ & Chakbot. can. ex ndeastan =f Ors —qypSters —¢ “provide relevent. amiurcss bcos — | provide: tele eng Nog oe ragenthnag_ Ssnewsledge -Stwctured 007) Po) ‘rep Hag. 0 Describe precticate log ic ft how it 1S attest +o Ke pregent Sietple_ Facts ce reletion= Ships in Ad. Cuait 3) a ‘Predicate legic 5) : oe ee I Preclicate [legic ie algo Enowon OD First ~ordenr lagic (ic a femal syste rn aged =n Ad ito _tepresént Facke and relanenchip bet? objects Tin a Precise and Shuctured 00 ~ att -ex tends ro positional | legic by indbreduciig sos ait fiers te prediicoses. in Components of. = pedicale Tegie= ‘ ‘) Predicates > These are inca “Fhak eae Propewh es of oRjecks or cetlefont hip. bet 0 objects - : a Form exam Ste: Bothers CRovin A340. 5 Rovi lend ALOY Ore protheltae OL = SS aiabl es “Symbols yrrat creind- “Pee vobsecks oc entne ia the domain. : \——— rele XY, % ore representing snes eop\e er (yee roars Specie name? oF So ieehe: raples: Rave ond Rory as Constants: niheso i veads od anti fen ( YW) - Eodicarey that’ SarpiAg | Te hue for au objects in adéma Ox = hfor ott , ever mits rp thio. pining , evewfone) je- x Cloves , Chocolate) mean : Pine: loves chocoloe. ' 7 i 7 - Wexten hal quant hes C€S)— SICA IeS Atbok Something is fue fox ax least one object ie 10 Fhe ‘ mon: syne x= (for Some, adleast Oe) Example- aX Loves = (#%y, He ecrearo)) means comeane joves Jee cream. =H Rlepresentsti on —: +t rate >" Ram is a Student”. [. Representation vStuden#C Rom) st i : ; : ' 2 Relakonclip= “Rang Ike ‘coePFe-” Eee aeeesien “Likes C Ram, coeffe.) 7 Guanes 3 VR moncx)9 drink Cx costhe) There are att xX where xX ig a mao7 Se Mou) A 2) NattS” Rction orienkd ®) Far oriented { boo! —_Lenocing “hoow) j@hat : sg 3 Compiext fi eaetee yy , je Sue ae af Can be complen, invelving Sim pier to represent 10-8 - _snany s SPS_ov algorith idle about Forts oF. 1 i pelations ki ps — ; a ~4) Hexbiliy 3 a) Pleaibilitys Less Hexible 1” Pore Aexidte. 8) Example iA ee ©) Sample in Aas A pathfinding algonthm(n) OB rule ia O0- expert that specifi how foGing system | "All humany whe Shoctest Mute - are Mest SLAP {vcathons in Aa! . ai plrcsHarli2 ALJ hecth in robots, oudiociormes Qsed in Enadledye —bosed s cies (lata :60 pl os roue Ss Ccsps) ¢ ood. _ discuss. “how, constolor getion can be _ ) to infer Soddihonaw _ eS ngyroiot Ss oCanid 2) froint. Satis Faction Problem CesP) consists oF anree Components Vy pec. - bles > 1, Xa, ---- X9 ~ the EEG LENS \ Vani we need to solve for: 4 Domaine — D1, Da, -~- »Do- possible Vvosues thot ‘each vanable can +alee. 3) Constraints C1,Co, —- CM=- reghictions on Yeombtinattons of Values that certain subsets 6f vaniables can tate . Example Ad exa rople oF asp is Gudoky ,ohere tne vahables ate rane cells of the qnd the domain is +he numbers J Apug h— 39, a0d ane Constraints Involve ensuno hak 9° ta ony row | dolores ae Search Gionthen Por odvantog’t 4. Hladeaniogs: the bacletracki ng Search algowthm is a basic meinod osed to Golve Constant Problem, A's like ing to RU in a porte Piece by. plece ;pubk Whether you place apiece that ego Ab, you qo bate c, and ty & Oi terent one, Saving CSP i =Se eH eects es “Sith anem aagignment= Bt the begining. e = RE the wenetils have been given ee a beose O Variable- Pick a vanoble Oks to assi¢ value.) | __ 9 Astign a val&t- Try assi ping oR the Waniables dermain pine set of poses 5 By pe thad doesn’ + Vin late any constraints: A) Mele “to the next variable = ohce a valve. ~—SSSigoed ta the current variable , move +O — Ne next one. ; 8) Badbteacke if Need ed. SF No valid Value cn »_ be assigned to the Gunent vanable vworthout ~—breGbNG a rule, unde the last assignment ~— ck ty St different valoe for previo variable ~8) Repeat = Peep aacigning values Te Vanobled i pose eeckihg when recessar + Advantages —s a) Sitoplicity = Me algorithm is Secught forwnd ~ to J andersttand cd implement: : —_complete:- 4F there is _a col? baciotrockirg a7Seasch will And j4. a+ ties every posible Combination, o it won't miss the /mght one 8) Space - eFA cient — t+'s doesn't require a 2 lot of memory: cae : 2 | __ ae Digadvantag e335 : ; : __ \ Slow For large paoblerns— Backtrashing Rad tne best woiv'ng problem together Single "pain to solution (AnD) € othess Can hove ‘| m att ple possible Loletory E (of) ch) Ex maple — ~ Cl €xample- Novdqation o< puale fi Gone strode or : Solving ; dearyien ~ mokingG18 \Onifosm Cos seasch “algentho—* (an) | iS tS cohith hag the ‘lowest cumal _| SF a5 daed fox “ravensing A —— edged Hee ao h oo PSS ae femed inte aio} when a different cost hvailable for each edge: Saisie eipee TSAth$ (goal oF uniform cout Seoech (0G) veasch) to Riad the path to sne_goal nese Fe : Tt lexpands +he pect nodes aceeschin 4! _to_theit path cost) fem the root osde._— =A UcS blgorithm is implemented by the T priotiyy Lee. : aa a4 lain Moarximury poontt “po “the Towert Curnutdcive cost r 34 is equvalent to BFS if the path cost of aul “the edges is the-same. Exam ple ! ey Goal Nod e—D : PAG ; pay ee [path +o goo} node - | A-C- DB = 3+4- 4 == lowest poy fo ~ fe ach qoal Node at every Gate iPagan ) wae]. 3.3 ][ A*® search algesthm—r Lon =3 i A® agasth®m iS widely used PodhFindin aad psaph traversal olgorith'm -fhat Finds +he™ Shottest: path Frerh_a “last Node TO good _ node in tweiqhted aroph. TAF | algonthnad OSes heanshe function hiGedee and cost fo reach: node: 'n! “Prony Stork cia ip)" = atgenth Finds Choertest: posh shrdug bh Sealech “Space - = Rkrmwa — : Feo) = Cn) + hen) whore Fen)- e8hmoted cast i ~4tn)— cost to teach nade hen) - heunshe value. aetromp le Da sto- S r - a + §-<—3ep} at sen 1a.{ 2 7 © © ? : | ~ geet [- Fenyehtn) q0n)_ | SPA = \4e@=F NS7B=4+9=6 S) SHADE = 14+8+l=F 8) SxB C= Seer n\ Si- ASCID= 1454+ 840 =49 = 6) 885 C5 D~ 4494 8F0=1g Ste 4) $7 Aa B83 05D =1F9 4043408 Shaan fo AAD = Itl2+o = 19 —9 fey © | SAbs8505 BD Aad Shorter Pah C38).