—=—=——-”~—‘( aw”:”*é‘<‘ ; Rréadth Finek Search ttadth inst Search © Explores ot! dhe node at given depth before proceeding to the next Jevel. © Queue Fs used for Teplementing ars. o Algoritym :) Enter starting node on Queue. PBF queue Fe emply chen retuen fol! g stop. ii) 5¢ Post element on queue iis goat node dthen telury Success & stop tise ly Remove and expand 4tngt element trom queue and place children at end of queue. Y) Go to Step 2. Exam ple . @) Soim= 0 “inPifotty Queve: Fa% FE ad 4 Remove From queue g& add ® © © ©) Successor bo qveue 3 © © quue~gcceg” SAMARAS € © quate §o.0.62 to Bnd oF queue. @ queue - Poe wry © queue- § EP Gif @ queue - = GHt @ queue - £4} Pan: ASBIc— D> ERP F454, © scanned with OKEN ScannerAns:— ° 4s} {a.9} fa,c wt Fc 1ds GNF F DiGi Ee? THe, ry fFiikg f © scanned with OKEN Scanner© Vasant of Des © ves hes tsue f endefinite Loop - On Depth MmPled Search we define a mis to go Sn depth, Aemes- © 4 fs sfmilar 4o DFs bub witha tor © Cond? © Foflure Value CThere Ps no SoluHon> “Termination © cat-ost PPlure ; Terminate@ on reachin predetermined depth . ‘There Re a soluion bur ffs nok Fourd afeven oper ae predetermined depth @© Advantage: Memory Effient. ' © Wisovarkages 0 Ge be lerminated ontthowk Frding solution even ie Jb exists . © Not opkmal. let Depth fs decided Example : (2) —rsading node —»level 0 depth d =2] — > level 4 —— \evel O This Level wil) { predetenine not be Kisited J imib- path: R>ASC>p> BIT © scanned with OKEN ScannerCase of : at gen: Node Ts ®) Here Depth 32 so even th u Ys existing owe con Nor appsy solution as there Bs predefined Wart of Depth=2 Heres No salulton —> cut of failure, © scanned with OKEN ScannerC—O Unitorm Cost Seorch Aigontnm: ® Used for welghied tree | weighted groph Troversel © Groat af she algoithm fe to find a path to deslfnalion node With Jowest cummulative ast fe. opHinal path . © Node expanfan Fs bosed on cost of she path . © Pafoutty Queve Rs used for Tmplemeting dhe Bator COst Gearch. } © High Pefoviby ts goer to path mectn minimum cosh. ¢O Bock trading Rs also applicable fn thts Atgostnm. @ ae optimal soluvon a@ minimum cost Ts conserd. ® Oreadvontage: Con steuck tn an Infinite Loop. Example : snlobe t= Difference between Queue & Prvory Queue? dn queue FIFO rule Rs used bub tetr Starting node- A in prioity Queve Highest ree Glool Node —G@ Clement 7s deleted Pare. —_ 12 Path ASsc4,9.6,, 1 — Ca & | © scanned with OKEN ScannerCase 2: If Goa Node fs H- TrFe lac node but hot goo! Node So back track . ® _, Prom € shere are to option after bactiractng from Gi te Dae amang which # 7s smodt ° 80 ° db Nod Geen Ge One Lene A-c-@-E-G W_, ash A-c-8-E-G:-E-f —, and Anc -8-£- G-E-p-€ —s grd abet A-¢-8-E-@- 6. 8 Ep oaath Ar node D there are shoo ophon ©4+ among Which c fe alread Vested | 3 1 Ane BEG ee Bap OY nod® Saurse Node, ~ © scanned with OKEN ScannerAiea Ge DEKH DEALS © St con be consUered a@ combinalion of DFS § BFS. O Beg depth tims Ps found by graduadly Srereesing mid- © Onother hond We cay day that we TnereaGe the depth Mnk- Sn each Preration 3P we ‘ve 48nd dhe soluHon goat node) — Turk dso eae raHON — Increase a © rot! *y Frocess : © gn Prat dteration d=o So We can see ony s. «€ 4st Jieration ,d=0 , Ls] rd Jkrotion, d=4,0 s5 age] a! i CSSA ard Gleralion d= 444= fe" Node ) 2, Lssasnoa pI ee & A™ Gtealion d- ost [ eomplete Traversal]. ce © scanned with OKEN Scanner© Recursive gorithm © sStat® Prom rook node and Poltowe each path +o its greatest: depth node before moving to the newt path . Im plemented usin: fame 10) Aigovithen: i) Enter Root node on stack. (Push) W Do unkd sack FE Not empay Ld Remove node (PoP) L tif Node = Goak step YD Push all children node fr stack, (Example : Star& Nodes ® AnfHolly, ko enter root node. on stock. ® for Rand Aish at wee 251s beox Wehove to toke on einer [51 on [EY wore direction for Potlowing c depfh. ® wetake te @ fop 8 ef, =} o © if é c egy © € |—> No child oF G fe 80 Je | c rao @® E also dont have child © scanned with OKEN ScannerDFS - Example Yor Exercise: : ® ifs leaf node so Backtaack to D ere sk will bocletrack. bom GD Ds visited SO again’ backtroct to B here @ fs vicikd b & Ps nok Vished Sot E will be removed iY ‘a —>+ *& Ps also visiled so Proceed Por adher depth. © pee OLA LO | premny. A3RID SUPE SC GFOH. a cS © scanned with OKEN Scanner. Exomple -DPs ay © ORE 3) - RI? Go ® nN e © ct + Dhas no child fv So backtrack Ee Bis vigited so . 4 Cc doathave ay child backhacd® [Te maa a ay q cal © scanned with OKEN ScannerUx t—t—te © Give an exomple of problem Dye eahfels breadtn Pret search would work beter dhan depth Firat Search. Give an example - io problem for chich depth finst Search ts better theo jae readth inst Search. S: Por the waler 4g problem BFS Works better +han DFS For there proving problem DFS \aloras better thon BFS. @® Descetbe how the branch and bound dechnique could be lsed ta fond the shartest olution to the water Ma Problem 2 Ans’, Branch and bound fs depth QRicat Search + example Like the ‘ater \u Soluion require on echnique. For problem there the optimat a small no-oF & and al sleps pe equal cost. oe fret earch eA better er | fe eee Qird the shorlest path + atthoug? we require a Neor. exhaushc Search ie Cone may be able. Joondd gome nodes at the final Level atter Sutionis un | will Breonch and bound will on. cot of paths chat are more expendive dthan the best ft found 60 dor. So we can Fmplement a depth Jingt search four it would be an +o clo so. St cowd bee Jok worse dhon a BRS Jt it explores some long path ear on. Branch and bound Tsused for probleme Hike fravelfrq Salesman en which nok ot moves have Same cost SO at some short poths are Worse shan some ther longer ones ang we would dike to prune them. 4 © scanned with OKEN Scanner—_ informed Search OHeurene Search @Heua dre Funcien ° find@ opirmal Galutton to reach qoak © Sr Isa Cearch which tie@ ho weduce omounkol search dhak- mug be done 4 making Pnieligent chotce@ for the nodes thak are celecked fo, entnatar O Heawde search) © Takea to solve problem fn minimum Gleps or cost ! © Teiea to oplimize a problem Uding heuristic function , © HewiGlte Pundion Hen): 3% funcvon Hen) hed giveG an eblimalé on cosk of Gane fem node n to the & Gtake . tohome d search :— fen) = Gen) + Ken) Fatbost HemteHe value. 0, eat 300 Gave) 7 &® a A\= ID +300=810 Ree 204250 = 250. O Two gyre of Heute Puncton: ® Admissible » HEM) < HC Goal) © Non Admissible: Heth) > wacgoedl) en: © scanned with OKEN Scanner4 dourte: tet & > Start Gi —rQ0a8 ans to heuisle Qunaton Ben) = Gen) +O) so oa “an € mle have ner Fen) = + &tN) we a PCe) = 14s 8 Fc) = Ifacs ales Fp) = 14526 80 seted @ Now considedng Nogram direct odual cost fromm A SG Ts lt 940+ 2 =1) a4 3 tho mean @ JtTsS lout pis Showin non pa} —> Admissible Letu@ concider sthere cotld be gnather poh from > ko RE UCD) =5 bub A—+D—> & —!_ +3 ex 1 : heeea (BY © @ Hees) u ciel Here HCH) PHCgoal) —> Non- Admissible. © scanned with OKEN Scanneree x _——————— Gree setentasesSenrshy Rik Search & Depsh Frat of node Search . © CombPrarion of Breadth Queue fe uGed to Ghore Cost © Pron! . fifa @® Priovily Queue tP@ containing tathal Glare ‘oP uy PQ = empl return Fail deny > eskirna ked cost else Nope + Remove -First¢ PQ) ip Node = goal reluen path from cinital to oak Node and and else iil OM Successor erated node Into neeac a PQ according Fo test Value. he “gad TA-Gorted-order) po anie are, cooHcl Jet ue conside Ms cwith Shralghr ne diStance re coordinated clistance- Given AG =40 Boh 282 wgpek. ec HE OB DOG = ar EV olG PG Bows d soy ‘ate have to maintal eee close land whatever node i 26 queues hen sepauea4 ome at dk in close queue) Rinish oe we will © scanned with OKEN Scanneropen Pao Cal sorted <—[68,D] order Cor8e,40J [e809 Suresor of = [p.€1 8,0) Ctr gorted manner Ce,ea.o] Successor oP — [6 .e, 8,0) dose cd CAT — removed CA 1c] —> Removed CAC] CAF “~ L meshed Qithe Fa, pare already visited £0 Cee pd fil tetrn, pal path —>» A—>G—9P — & Node HIND Oren close Dram te Cs3 cl _ Cel ts c— F ae : e— 3 Cee, (3.83 eee ae Saar CouT,€,03 Tse 4 7 Csered | aay ——_==4 J I— 04 Poth + s—>B— >f-> & FT} works oF Pun) shen) white A* @ores with Pew) gind shy © scanned with OKEN Scanner= — — / AY Search Ox voce Preuttire Raat hen) and cog} to reach nde dhe GroteG path through search Space ist and oflfma ryutt ® Algoiithm: Enter starting nede Th OPEN disk » i OPEN “ied fs celusn PAIL. ii) Seled- Nove from OPEN is which ha Smalies} value (qth) Dif node = Goal ,retuen duccess WW) Expand nate ‘n’ and generate aul Muctessae 4 Gmpure (gth) fox each tuecess6r node, YD if node n fs already Yo OPEN / closed atch ts back pointer vi) bro Fo gth =1462F Boo S389 gra arene cok D390 SB = 449H -L K>6 fai) $+ BI¢ 7D = 449 434029 ene. Re S>A>B m- SAB ~ l4eroe 5 Suond [SA > DS les O=T SAO legate nae = 06 ~—l|sAve] [sad _ aplo 5 x z shBc—3D this bee a web) CERTEAYAS = 08] o[5 23] Ms ran A © scanned with OKEN Scannerfen = gend + en) sl ee > two opliong. > te = G+i4gal4 Direction © Direchon oF B S>€ S44 s>3516 SC will be Considered as sB—>f | SB-—>e one SC >e ;Sc—d = = 4451) 4412 $4- PStHot4a | stigrg ; gen) Hen)! : > 8+ > SHo044 toy = | > ‘ & j ! explore thig ., Sd — ve SEPT ARIE | BN). hen) Now Scde—y@ 34% f24549 aly wn “cn) © scanned with OKEN Scannereet witb! SOK su I4a=4 1o +0 =G|RI SA>B SRC SAB 314244 =H f SA 3 Ie 2 = 04 SAC— 9D 3144346 =U SKC 3 Ga V414A40 =06 Kes snortet parkh SrA 2c = 06 SAB—> D-A5 1 424546 = 14 SACD > & =l+1434 240 = OF Rg SABD—5G =l42+ S+2+0 = 10] © scanned with OKEN ScannerLocal Seantitoa A othr ocal Search Ajyonthme: OA local Kearch al gosithm operate using @ Gingle cument node Father dhan mutts le path and generally move ondy to Neighbours oo ie eS © dn radii 40 Pinding goale , dhe lowal search al orithms ore, nett for ssolvtn pure’ oplimizalion problems, Pn Which the am Fs find the bea diate ecording to an objective funchon, Oi-e. Path cost doew not moter and on pomug on scution arate Needed to reach qo Node. } Advantages: wlocal gearch al thm use a LHe or eonstank amount oF as ag 5 opera ondy on a Single path. 5 lente atti © Lecat search algorthrng are asus Put for ssteing pure aplimizalfan frelolem sin afiich aim PG to find best date ee objeckve Panctfon . © Slate - Space_londscape \ © scanned with OKEN ScannerGenerale &, Teck | rasa ‘Soluon, Tedkerr Correct nore STOP © scanned with OKEN ScannerUYU CLimbing ‘Simple HIM clfmbing : ° Ngonthm: 4.’ Rialuate cthe Triltat state ; & TPH ts also a goal state shen rebuep at and quit otherwise Conte with Pate stole a@ the urrenk State. 2. Loop ‘unHit'a saluifan fe ound or aphh there are Do new OperatarS ett to be appied dn the Curent. aiate.- - : 40 the current atate and a uk ta predue ANewW new state bb Evaluate the New state D IPM Ps a@ goal hate then reluen @ quit » Rt is nok a goa! Stale bub TPs better than She curent State then matce it current slate w» df tt ts noy beter than wrrent Qui} Curent aie = nial state yar Sperator '0) and gt new rate (Ns) sol?—> Qui CS=NS Example: 8 puzzd Problem. Vj? 14 ‘ja +]. ; Ss | Selart state ole tH gem stale. ants els © freblem- fn Hill efobing /Ginple @ steepest} LOR ob © Loca Maximum: a © Platue /Flat Maximum : (value oF ott nefhbours yi \ are same) ° . x z= - | Ridge: la ton) change chirecton . 4. © scanned with OKEN ScannerAen Steepest HILL © Fito ‘i 4 Difference benoeen tinple G Sheepest Pall sdtoltng * In sheepast hill clfenlaing MulHple point@ are chected , dn Aimple hill climbing fiat state which 7s belter than Curent dhate Te aaa) od rest of the lale@ are nok explored Wherea@ fy oe Ascent hilt climbing Aigo selects the bex- ‘among sthe chil rep atate@ dhabrare I chan the current {lole. | Lat Move@ are considered and beitone 16 selected a nent ° atdbe. L Examine@ all néighbour® Node@ and eded nlosast to sold OS Neat node. Evaluake Entfal State : Aigetie) Bt 3 Bs alia cal nade then sebugn Fr and quit Otherwise continue With Priva stateo@ current atale . Diop Untid.a solution % Pound or unkig a complete Treralion Produce@ no shange fo wurrent grate: E D Let suce be astate Auch that- any, possible successor oF Werenk state will be beten than sucte ® for each operator and gonerate New state } Apply an Operator & generate ne@ state D Boaloke eo dole. 5) BIE a goo) Siale shen a bi se WWE BE Nk combdre Pe to Succ e NOt better Jeave en Sef Suce fo HhFe state 2p D BF Succ i beer dhat current alate sthen Get curren} Bate to SEC, Note: Best plate Ps node with better dora Penstette value, © scanned with OKEN Scanner\Pmulated Apnedting # Also vatalion of Hill eltmbimg 4 Adlow€ downword Gep At FE dependent On d Meballargy theory where Annealing Ps a Precess dh Which metalé are sdowky Cooled te mak a M redch & stoke of Pow energy where ahey are aay Advontag: Disadvantages Simulated # Gaay to code ¥ Slow proces ~Annealin tor complex x Thoonk kel! +) Annealt ob ¥) Annealin probler cahether opt a Damen 2 ls Gioe® good 4olulon Ps found maintarn Solution Ly For which odher| |.» mous to a Slacisti method Pe required worst stotel fossstens tid nedpred opal gain aad AW) Beat etate und Ts als motnkained © scanned with OKEN Scanner