AI Unit 1

Download as pdf
Download as pdf
You are on page 1of 21
[eats + “ifieal Intelligence :— AT Technology. EnalolesS computers and System] “to clevive mreantrr pub Fa fornato from la ital enaqes » Video's ond otter Uaual Pnpuls and paced on -those. Yoputs , St Coy “take. agion. # Tt Go the Science Yy mnaving machines | that? Com AWK, Like humans - EA Sp Ca] defines * ean —made» # Patelligence debines™ inking, power". | +50 AT Mean's rnan-made thinking. power: It 25 a lovanc) S\- Comp. SHence. bo voli chy we lean, eveate. Siatell’g ent roochines which Com behave tke a human think Like humans €& able to make deasions- ea suo Robot Te portant: — fa Sho Health Tssues | Take TEs | Vidual asttstant — Google Adistant oe ‘ Tp Recveste tape _Sye Date. ra * Tangle meat Homay Trtetli gence ‘4 machines Contibuile > Computer Science = Rearoning > Siology at eoxning Solvi oy 2 Psychology, > problery beluling i > Lin ic8 2 mathematics | > Engin ceving. Fro qramy woittaost AT & with AL > -Arhuer t Specie PGeneric > rredficatiog 2, oot =r modpcatey A Simple sth & Easy A dyenteg es & High AcCucacy, woth lewd Evrors * High Speed a tah, Retiability * use fal for Risk areas. *® Digita| Asc Slant we useful as public UGLY telecine hat “lan Enfroament usher the State & mapping too | ‘L dested outcomes - | degined ond Use ot on atomic Level citthout | oy Internal Stale VEble wth a | Coheng algostthrn » the prololen~coluty agent | IU eaten pisadvantages/ : aabgh Cost Cay not think, oot Qp tWhe bon. x No feeling % Emptiong - | & ID erease, cle pen dence X ND Wi dingl cveabvity, On mac lones. dblem Sching. tun AL. The reles agent Ok A clirectly. maps states cto adtien. Whenever these agents fal to operate i) | Josge. and act easily. pugormed boy the agents “hey the Sorted problem Aiscolues and cestt to a problem: Selina. demeain volch breaks “the large tered problem Foto SFentlo Smaller Storage Gea and ve Solves one by ene. the frol Sotegrated action will be the On the basis ye the paste and Aner ssorking| clometn, br erent TYPES Ole problem. gduta ogent| problem = Dae i I | Lechnrques Peron precksely by dle bien prololems ~ uy | a Seyeval goletions. So we cy Sou Hhok proWle Gding. “a a part Oe AT thetffrmpasses 0 M0 & | | Sock ag= twee, R-Wee hevogstic. tnms to cleo Solve a problern. protien —Selsing agent “is a Tanovalle ;— Th which Sdutiorm Steps Con be Tgnoved « | —-Y Retoverdele :— TH wolfchy Sdution Steps Gn be on dow, “7 Trecoveralele:— Solio steps Cannot be undo. Steps provley Solving. in AO he probleny oh AT Ws diced onotiated Sth the nature Ge homans & hee activates. So, 10:Js- 4Prite steps to Salve a problem wolich mokes homey a vwoorkS. Requtred steps’ _ provalery Aebinitiog :— petalad Legt: Acte 4 powiel< = IG pate. os WO UN certo®y tn formation. i|moke dec¥efons based on -the Spectkic ot eee ener s Characterghics Roblem ni | Al proole reg ve fo th o wide ronge a tooks | Ahat Gnvolve the develo pryert Ble dlagedlnens (or) Systern i () Hhat Coy PArforen tasks that typically, Cequite Kune | ri } (evel “voll gence. 9 Suck as decision-rnaking, Jengoge | : F 4 PrOcesiNg- InoGe veCognition ,ond problem — Solving « | &Compleaity I= AI problerns ore, betey highly Cornpletes the processin ie forge, omount Oly dot? and the abei Gey 40 hon oncertornty ond ambiguity . The algovithms ond Syrtens developed to Solve these problerns must be oble to manage cornplestity. i requiring Grd make deci’stons bosed © incomplete | Nen— Fnearity i Many. “AE prblems éahibit Men \tneor Relationships I meantyg -that Small changes fo the Input cy lead to Sfanthiant change Qn the Ostput This Mmoakes tH life Ot to develom algorithms Mnot Gy octurotely pre art outtormes (od make decidoys based on Snput dato. * Context dependence :— AL Problem spten Yequive the ability to Ved es ton of ree | ae ©) egRroament- bxte 0 Janquogs. preceiing Spier cejagt be: Ute able 40 Unmeler Stand ~the rearing Jy L Pothe Content Gy a Semlence Cod poregrer? 5 rue + th : K Creat sty s— Some AN probleny vequiwe ¢. | ws : : » ideo, eae he ‘to 6 enevale, novel ad shioms (or * AOd, i Bytey Using techniques Such os energy & advercertatl Network (61 Anis) (0) Evolusionaty. algorths Thee fechatques allan AE Mptord +o | team from Entutting. clola% Generale nu & tnwnvotive Solutions to Carqplen problems. Fleet Alea’ adap tale ee AL problerns deter require the ality. to lean fo adopt overtime, bated Bo Nw aato (or) feedback waht 2 ochi ened) throu! techniques Suh as machine Lewntng , deep leaning and reFrboenent learatn juahtchy oable AL fetes 4p Ura prove Wer peabormante overtime. * Mutli — lis cf pl?n org 3 AL problems aten vequie. Expetise puro gputvole folds Wnclucking cornputa- Suerte’, ercconbut rnathemati cS 5 Caahtics & tongue pry chology» Success pu AS feptern roust be deigned| and developed | hay. Acorns bith o wtde vange Gy SKU Knowsley Constclevaliong:— Al problems yatee ” Secucty iFolos > fakmess. the. develope’ dy “Al gihteal and Socal Concerns retoted Lo pvaty. » Gates, vequives Crckul constderdlion jy these Mua ty Enture hot? the. ‘lechnology 2, used ethically. & responsibly : i | Search Techniques — | Re Search technique %s a mechanium through which one coy bind relevant -oferendbicin been || dakormation Sqpterns the. wlrormmation Suystern mag be 45 npuse(od online . i Storage mmedqaqism Con be Lictionary 99 mete | S chawikied.. Owrangement Sf xo kormnabon. | % ASeorch probalern Consists 6b- +A Stote Spoce ¢ Set Gf all possible states eohere | gow can be- i >A start State ? the. state form where the Search | begins ’ | A Goal State tA function Ahat Looks at the Current | State. coetuans whethercen not it%s the cal state 4 > He solution, to Seovch probleny “by a Seqrence MK | the, stort) actions, Called the Plan, thot trans forms te {@ the goal State. i Paye No Nie eons = x ahs plan “us achieved Mhrovgh Seay ch nl goritnens. Types Q- Search eae Unikormned Seaach To formed Seas, chy \ | Bnd Search) Croesticy atatreationa | e Probléiny Reps ra Reatutiy end Sat Cost at Sear Search sexy Seuck | | Depth United — sherative et ast HEN. Cem bin | Seach © egga seth aaah + | (Uridormed [Blind Search — . oma koosledg e Such as- *& Tt doesnot Cot any a | clossness ane location g- -the Goal : - Ip opecites In a alle Fete wong ras at enl Anetudes raformation about howd to trodtyse the. tree, how +o Ideotihy. aks R geal nodes: 2PM follows bate force way . ; 2 Tt apple a ways to which Seavch Wee | "S Searched uwrthoat ony An formation aboet the | See ory Spoce Ble Initial state operator G test WL. the geal, 50,14 is called bind Seach. Page No Fajorned Seavey ary Ses cloroain, Knowledge, he probleer, Yahoo ation is, awvciloble, ulbich Con quae “the Search | Ly Dppormed| Search Shale gies, con find a Solition | more Fhicerst | ‘than a © | YY onto | 3 Tt ta also ¢ oe | | Med * peuatshY. Seorch 1 | 2 SN a a Looe which roighit" oot alvoauys be quovertecd ae Voeat whiten bit guorantead to find 0 pod | solution Ln ¢ eotOnable time. > Tt corn golve read lorrple prdolems ushtclh cald | not bye Sdlued| in another way. | | Breed 4h Ficst Search :— | Tt is on algotthm jet taverding. (o>) | Searcking t6ee (ov) gor Aata Shucture Te stots at the tree ~wot [or Somme anbotbany, node Shagraphs j | Sometimes referred to aS a "Search Key') 5 and tuplores all db the Metghbor Nodes at the Present depth prior to waving en to the nodes All the neat depth level. Tt 24 Fenplemented USA queue Ae x See select’ shortest palh Use ob d = R. —» Bt tira to follow FIFO Rule. 479 plemadal CQueMe. Dass an } Erat— Gs 2 Rae ® Solution:— BFS traverses. the tee “ sho llocwes node Bist") ot Loould alvoays pick the chablasest branch unbl 2b reaches the ¢ltion - Chorus, out Js nodes, gees tothe neil branch — level-0 [evel-) lesel-L — [evel-% or Levet-y Patty cea. Jp BFS = S386 S =the depth JL the Shallowedt Solution N= md. DE nodes T5 level 2 & tse pleat & aera to the ce. 6 Noses Cea. wo BES) ORG] the shallovsest” Solution, ? Page Ni a Bana “Ten) © Lente nd... an’ = 0 (ni) i space foment i Equivalent te how lorge | can the frtnge get S(4) = O(aS). i coropleteness BFS “os complete, meeting. fer 0 geen Search tree, BFS Will Come OF oo a golution 2h St Esdiatsy } Optimality f—} gpsian optimal as long ag the Costs! & all Edges ove Equal - De sth First Seowcly: oes of algpétthm ipo traversiog— (2) Searching tee (09 graphdata Sructuve. | The algorthm atavts ot the woot node 4 (selecting. come anxbitory, node ag the woot | node fin the case Spo graph) f Explores | as for of possible. aleng Eacl, branch be pre. beck trocking- Gt uses Lost Sin — fist out (et F0] shategy. | onieg hence at “no imp le mented sing Q Arock + s Tt used +o Select lon gest pails wen bee 6. @b 0 9. ee ee Solution :— ole faudualent Geach chee. toe the above graph 25 ag follows. AS PFS chavevses he tree © deepest node bet? ot Laoele aus prck the deaper branch, vst| Tt reaches the felubiory( Pt wos out Ss nodes, & goes to the next branch ) See Lee oe A — level 2. @® oka é © eau bs Page No, path wood dy Hrs: ¢ = Na b> o> 4 | d =the degth & the Search tees the no Bk levels Qk the Search tee- n= no Ne modes 95 Level % | pe Geeerplenit yp o— Eautualent tothe nod A nodes teauesed In BS. | Ton) = jantantyg, . and O (m4) Spoce Compleaiity 2— Bauiualest to how loge, Coyne “ogc get . S(p) = O (xa). Completeness {— BFS 1 terpplete Sb the Search tree 2s Finite smearing for the pen finite. Search tee, BFS ull Come, op witha Golition Sf ot entists. Opti mality, s— BFS % not opicnal , | Meaning the no- Se Steps “ory reaching | The getutttery jlo the Cost Spegt Py reacling ato ky Wig hs | oan eee ont suclod : Sydterg + —— } pro’ ie oo: { Roeduct orn Syctesy us a Powe work thak avdsts q te | fa developing Computer Proguarn’ to automate a wide range G, fasks . Soe Stans Wi cowsllas Genpacts the Creabtin Ny PI-based A 1absens Sysens Bke Computer. sflecave + cadile applitaten: and Manu Weaceeeod 4vols» DT’ Saves asa cogettive avcl tecture. VITt Encompases vole representing de cloaatve | knot 9 allevoing machines to make ecieiony and act based om ck brent Condittens- + mony. Expert fenos ond, automation met ecologies. rely po tre les clegrned mq °xt- Revel ealge, Bose Ral Tecteat) Bose i z <-> top / Tnerence—> Active output Engine meres (used) # Ws avclilectte, wiles Gtictured ay LHS Go RHS equabOons- 24 . () LAS > Gepeuke the Wnaetition to be foatvated! QMS - deter aes -Hhe ote (09) get moutiog WOM) The Egtanated! Conatiting . { - = e, £ ye PRRAle Based appeach Coveng, Hye Py ary dation : De Boetuction Supt ern, ® Representation Aves knowl edge. Ca A Goenpvise, Veesous Components vsog §ov snaking, Seoteltgent” prachines - Components :— a. arckitecture Oy & procluctiion Gyten | a ALT Conceats ve production ruler sa lata lee ress tystero. Conbho [contol Syiten) Syatery [contol Syiten) ete) Vv | Knowledge, Glolea) Bectaloae | ee wa0rkk ew | -7 Gloloal Wataloase .— % Used aka Cental date shucture . % A claddloare ComtArs atl He necerar data and “ukermation requSved por -the Succes ul comple. Ae a tasks . Te com be Atided Goto tin pots ax pamanent &p Care permanent > CNRS He Breet actions Temporatg > alters oClovdtng, ts CTreumn stared. Rodudioy Rules: — the Set x e BF cwileg hat’ opevate on the lata gethed fRLern “the Yobat clottalase - these ave bound udith precondition € post Condibopp st ats checked “Whe, clataloase - KTE a Condition Bs pacced tworgh % producto vale & get Sakae by the qobal Aatobose , hen the wile Ss tuce emfpelly. apeted Conteol_ System — Se checks ‘the appicltlty Of sed) eo helps decile. which, cule ghoul be apeS and tearcinates the procey woke “the “—_— cowed output « scoxqlves the con a wule. aves the , . yet oy coultirle * Tt olvo Cen diBBOS arnating of the Sarre -ame. He strategy. Js the conheh Myton Species’ He wld Took compores He PS) Ane Gequend do wendy ; " pen ahe qaeoroat Aoleloose, tne torvect F : Chavoeer® she mrad 3.0 1 aty 3 wie a step FO cateroert ever ave) BIR t ect agen Se, Dpre oledge A Helps we present Knocoledg 0 Golve Veal © . are poduct® ordearclands,,24 qt ia ee) : ne ve a - | Aelps Smpeve the veodolal tye “y 1 d- the production ule. | | date. F —— | & Modularit i v8 ay the x T's lacrernenteal Aen povernent” Production mule G4 be Qn cincvete. poss - : de 9a formate | SK Tt mode fen collecten ‘ a dependencies | Rp facts rock mong. not hove GOum tes 2, a wide connec together. AL The. ad lod deletion Ie Gingle An foren OD I ertl oot Wave mage Eksect om he odd. # Helps Enhonce the performance Op Hee | PS ly adjuding the parometers dy the neler, | Modikialat oo : -& helps atte the mite o per weQurramenctks . & okeletal fen PLAC Created . % gether The requtvernewts G make Chong Raq Ane vos Shuctre Ue the Ps. Fe Skerakt ue. Snprovenet Oe the Ps. kno uoledog — “intensive oe Bente knowleclae Pan foe Deo humey | Seeley lan as Ea. | & T's .nok ae usa Ong programing amg pe DBE needs plo Sen eee i & TE whale. Come turd ay fen hero throeye Sentences.

You might also like