PSESSSSSSSSISSSLLVLLIVITIIITIIT IIa — Bondi inst Danek, Seanch _ Ly At scanch Depth Piast counch teptt limited — Geanch AD® A\ponrthm Tate Dep dep ch Fast Sarnch v Preble Reduction Bidineedinal: Soanch Hil) ¢ hoabing. Univonmel | Blind Seanch!- Tk doowt contain ary. dornaiin kwouledg ¢ such ar” closeness, location of goal Tt oponttes im Rnute fance aay, as it only. includes info. qhat how 4 traverse the tne 2 Imw do identify — len? + goal nodes. i nfermed. Sounch:- FT “uses domain knowl e the problem inf is avaiable Which can guide the Soanch . a? A heunishe 4 wey which might not always be gueranfeed fn best soln byt Gevrnteed ty find 4 good. soln is mansonyble Hine Brewdth inst Sanch (8F8):— “Syovhy Ane Jovet=o FIFo Qveve ' ; Tine Complexity = b+b inst level=t b« re bench. ~ A leved=2 d> dept Lagan. g e ey Jevelas Space :- 0(44) lercletranma “Aly Kare AI nefrs to simulation sf homn intelligence in machines thet ane propummed to think like humans @ mimic thein aching, is, _(soering Type far ee Health one coe an Automobile Weak /: Sinan Suprinttlhgence - feng" Nannew Lommponents” of Ais + Applications s— Taye Recegiton © Types of Modelgs— Speech Recaiton Deap lessining Chatbots Machine leanmng Natural language hreneretion Sentiment Analysis © shu &Wlw for +tnaining &£ running modelo? — GPUs, panallel pnocesaing toels( ke spank) vchud data stone £ comput pletinome . frogramning languages — fyhen, “TentonFlo Tes , c entiiid eadasnap wii amigy L Resiekve Machine => These machines are mort basic fines of AT ailing | - Desp Ble, TeH's ches-Plying Suprconpaln. AT teams do not use any frainina sets to feed machines non do kta shoe date An fidne efnences Based on move made by opponent, the machine decides rect move. Neutnel NetworksSESSSSSSIFIFIVILELAAILILLI LLL das = —_ ee Dinette DeeperigDh: ~ Bs omplnetion of Fseers . This alye pornms depth -finst seach up to a contin “depth innit” 4 it Keeps intnensing the depth fimit a}itr each jdenation until the gr! node it bond « Adv.> This a\o: com bines the benef of BRS's fast Sench 2 pec's memory efftiency Dis> TH neperts all wonk 4 pneviows phase, Jeelo 20 TA bL=! Tas A,B,C ‘ . Aevel £ Blea Th ADEE, Newel 2 ores Ty: 8,0 H,2,6,0 fF, KG 4 Jevel S +n tev mQ . Ba dbieetiondebeeseanelr=: — Tt doce Frnwand soanch # backwand Seach + Tt uses BES, DES DLS ee» > One chat fran sur node Adv. a fast Jess memony- and othortnigon) node + Dis. Taplementehion is dfalk» 7 Stop. when intersect. 7 One should know the goal site in advance, 5 feet node Brwerd Backowwrd AY Z : - No Sanch}_Sorch () SO on Ov% ' We - ye infensechog = ~~ 6 © Node Gog nderrr err ere eer PSFK KEKE SEEK EHTHEKE SOY Alpha :- The dest (highest value) choice we hove frond sofan at any print along the path of marimizer. (ns pisyen) The — jnitial value of alpha ig —o-. Beta The best (lowest valve) f minimizer: +90, (Min phy, — Te nermows all the neder which ane net really * affcting the final decision « et @ ’ Its rostly used fin game plying in AT. Such a Chess, hie-the-toe, go, checkers ete. —> Thee ane 2 phvens max & MIN. — Max fon maximized values MIN £n rninivized value . minimax perforros a DFS agonithen Mini- Mas !— Limitation :- Cornplete - Yes Tt is slow fon oni ion) = Yes games Such as chess go Ch Tins Coll O(a") (oz these have branching ton Gaphsts’=6tbm) So this limitation of minimex Spe Conplesily im can be improved fron FFRRBSPISISIIIIIIIIIAIIIII IIwee oe www vvvwwvwwvwvuvuvyw = -et3errrre T+ 1s) most ene AWNwWMHM TONY) RS paaaneeragaenemtioo EO ine seach. den from stab ste + h(n) + cost to ho the node ‘Yo “7 uses (n) Cos neae’ e ik best Fist Search» y oy ke combined fastunes of UGS 2 greedy ewe ode 9004 KEY) Jot} Gost tenerch Fem a to goal rede n thom Shnt shite Stp-D Place stenting ode jin Open lick, SHr-O cheek if open list ig eropty: : Stp-9 eet node from open list which Rag srr value of evgloation finction g+h ) Stp-@ Expand A west genvuate all +f tit on int> Closed ist. SHp-@) Tf rede nt alnedy in open & it should be aached back pointer lowest (mn!) valoe. Ste-Q Return to sHp-@ It successong & closed Wire, then which . netlects oe Sea ” SoD a? oN NRO ots Ineal [5,53 : Frewtin-t= 7(S-> a aif soos AD] tie [(s208 1), (Se, 4) (5-56 1093 Ty: f(Srnacs ail) 1(S4n4G9G,6), (596,101g fS3A498,9) 9 $906,@QPee ee ea sss sss s sess sscddiddddddoe _ Knowledge (ase Re pnestalation a Redoning => Reedlgemesid yestimy 11 tose ayerds whe have Capability f maintiaing an inferml chle o knowledge , yteagon = ovet that knowledge j update thein know ledge afte ohcenvetions and take actions. knowledge Base ic nequined fon updating krewledge fort an ggent do learn =~ with — expeuences and fake acto4 as pet the knowledge Typee_ of ome > bc as — tf ce ee Cerow shod Some tine) objet teams) “lee Heunish’c ~— Role ff ; a \ knowhdy e : Shines cedure” Prrwioos Peptrucines : Meta oenif empnl Rue’y Keautedye 6 ores) the, Proredune Nk ab todo Semething) trereseges es,RurchasejtheNotes Ia set Most{Questions) Per semester (ALL subjects) Tat amount will go UES ae og For specific UPI ID - Subject - 50% sahilkagyan337@yblEr Sahil ka Gyan 4) Stepsifongetting/NONES and) MostiQuestions;- S_Do)paymentiusing{UPI IDE efelanlle-Vopiecvekeeysennrgo)| and send me on Enefll - Per aCe Than finely access alll Notes and most questions ¢Scan & Pay Using PhonePe App SAHIL KHAN