~Types_of _ Search __ Alggrithn Search Inform Uninforen ZA | 2 Hy at Graph Seasch| depth uniform Aree heuriclic a ape breath cost || Geareh_—Search heuristic search first Search seach Ge (DFS) (BFS) (UC, Uninformed Searching Algorithea:- the _ seavch__algovithm — have__no additional information __on__ the _goal_node. other than one _ provided in __the problem defination. The _plans_t0__veach the goal state from __ the __stavt_ state | differ only by the __ovder __ov __ length of the actions. Uninformed search 5 also _calleol__ blind __seavch. Time — Complexity:-_- Te time complexity of a. search algorithm isan ___ expression for __ the the worst = case___amount_of time it will take to yun, expressed in fensuber, i p40 "branching 1 factor. _ . Time “complet | is " || Measure _ by : —— _ Ea o(b)) Se Space Complexity: eee F | The space complexity _ of a Seare | algorither _is__an__ expression _for__the worst- Case __ amount of memory __that algorithm will _ use . Space complexity is also aa by ee : 0 (b)4 7 Comp lete im | A search algorithm is, Complete i} if, whenever at least 29 solution __|l exist algorithm is Uc lied a. Solution __within amount time. ~ Opting A seach _alovithm —6_ a Lit fina’ a a stlution __which closest to _ best Solution. ee 2g Per eeBFs Breath First Search: isthe most common —_ search strategy for__travessing a tree or This algorithm searches _breathwise in graph . t oO breath - first - Search, tree or graph . So it is called || BFS algovithm start searching fromthe root___node of ~—the_tree and expand all__successor node at the current level before moving to the node of the _next__ level. > |The _breath- first-seaveh _algosithm is an | example of general -graph search _algovith Pe > | BFS implemented using _FIFO__queue dato. _stvucture. 70S: > || BS will provide a solution if any Solution _ exists - > _|LIE there are more than one solutions | | by a. given __problem , the BFS _will ptovide the minimal _ solution _ which t Yequire the ___least__number of sleps [! each } o Cons- Tk require ¢ lots of _ memory _ since level of tree _—must_be caved |~ {into _memony _tp_expan® PIBFS needs _lots_of _' time tt fom _the| = | solution is is ony fy oe ror e IN toot __ node. S Mm. level L (B by 5) ——level- 42 wots s | at te. ae ie or : | wy tevel—3 : mde evel 4 4. A es . BCD Kum gDEF Th pera H| Lin ~FGHT WN [EGHITK . QHITK] : (byt ALIKL oat Pak (3) - 84 L7kum n—_conplty is ibSNe] for BFS Opimal path is “6 A207 82G Time _comp lexity_= 2)" = 4 For DFS optimal path _ is: Soh 2B2C2G Solution is comple _and optimal. aExample 3: © a eel Oe ® Q level 4 —_ 6 5 a ‘5 level 2 i © 7 level 3 | | | pode7 Devs First _ Search: | Depth first Search isa _vecursive algorithms for traversing a Wee of graph data. ctucture. Tk is called depth - first- seach because it starts from the _ root _node __and follows each path to __ it’s greatest depth node _belove moving to __the_ next node. DFs uses a stack data structure for ___its implemen bation. ree The purpose of the DFS __ algorithm _is Gmilox to _the__BFS_algoxithm. | y Root node» left node —» Right node Fros- : —» | DFS vequire very __ less memory _as it _only fesds | toi ciovesda | ptacks OF the _nodes__on_the__path frorn_yoot | node to _the__curvent * node. i Tk takes _less time __ to _yeach _ the goal node _ than__ BFS. _algpvithm. NS:- ene There js the possibilty that __many states keep re- occuring _and_ther————— = th impo in solution. re pepo —> || DFS algorithm apes it roy | searching and $0 meotime 0 Tine the infinite _loo Complexity :- Tbe ati is Fah - lm lime. ee of _ DES will _°e et equivalent e 7 O(b). by the node _ traveresed the. ae is Space Complenity:- gin DFS algorithm nee to be glove only Single path from _the root__node_» hencell _ space complexity of DFS _is equivalent to the size of finding set. Tt bs given__by 0 (b)4. —— DFS search _algovithm _ is complete letness:- within finite ___ state space_as it will expand every node within a limited | 0 search ‘tree. imal :- DFS__algorithm ix pot oplinal it may generale a fay : as - steps or high _ cost Se turer _ | tothe goal Node. Yeach || = I~toot —tight - left As CaG>Fo Bs Eo) Sometimes, DFS can be incomplete - Use BFS:- When _you_want to find the _shorte ath from af ___ certain source __to Oo cextain __ destination: t to exhaust _all When __you__Wa0 0 u possibilities and___check__whieh one _is _l| the best / count the number _of alll possible ways.— || a 7 thm:- |. Algori Anferted Settee ration —_| Tre. _algorithons "helps _io on __ the ca}__stale , = ia waaaanaene Th p ymati move__effecient earching - fs_ informa lled___ | is obtained by __somethiog. oe TT Phewictic: (2? Heusistic Function:- Than informed —_seavch » 4 heuristic isa function —_that__ estimate how fre the goal state. close a __ state _is_b (Lesser the distance , closer the distance) Different heuristics ave used __in different __inforened algorithms. o- Essentially a _heuristic function _ helps algorithm _ to make __ the bast decision faster and more efficiency. t t First Search Algotithm, BFS is __intotmed search Uses queue let “Open” be a Pulovity queue —_| | containing initial state. Biovity will be accovelin valu. Loop, if open Ilo heutic tie Is ——2mpty thenreturn feikove : Node e— Remove first (open) If - node is goal than _yveturn the path from initial _t» _ node Else, generate all successor of node anol put _newily generated node _ into open: on ___ the _ basis of _ referels. End node. Steps When can perform the_whole - task __in the form of steps. Tese. sleps are__as follows: Place the starting node __into_ the open list. Ifthe __open lst__is__empty , stop and return __ failure Remove the node _n, from __the open list which has the __ lowest. value of (him) __| and places it __in__the closed _ list - | iv Expand the __node_n__and. generale the successor _ of _ node. 9 v Check each successoY. of node, n anol finol __ wheather _any node _ i IL node pall a __goal node. If any __suaessov_ nog __ Is node, then returns [____||_ Success — Mi || For each succe ssor» nate | checks oy _ evaluation. __ functi hh nd then check of the _ ee an. ; » closed list. IL been ia __eithey__open 2 7 wi | Return to step __2- sh Time Complexity: - ~ a CL time ___comp lexity oF best — fint — search _is o(b)4 Space _Complexity:- s. The worst case space complexity oF _||bfs__is ob). where d is the Maximu depth__of the search space. Complete i Greedy BFS is ako incomplete , even if the given _ stale space__is Gfnite. Optiroa.!- : : Greedy BFS _is_not__ optimal. arnple:- ; 1 Start _node= A | ss Goal nde Ge |- ol | Rigs Snake a age 4Straight line _difference:- A>G= 40 Bo Gs: 32 C5 Ge. 25 Dogs 35 E>G- 14 FoG: 1 H4G@~- 10 24> 6 Open Close J L) (g, Bd) {al [F, £8, 0) [Ay ¢] ‘Gy EB, DJ A, CF} A,G Fr 4 Path is Pa C9 FGAlso solve _by another eee A \ Q _ eee method — aL tcl ee O Iisa eal T — i [> joo IO | i ind si om A la FaG | le % B® Search Algosithm:. ~ AR Search is one of the best and popular technique The most important path finding and graph advant Ar earch algorithm he ee it from _ other traversal Separales ed Used in braver sing :: i TAA orithm Steps. - Fist add the = beginning node tothe open —_ list. C Then __vepeat the following _ steps: Now. you can save the path and + wotk backwards Starting from the square fom each Square you go, till it takes to the starting Square. You will found Your __ path now: 4 i Wh Yas Ak Search Algovithm preferred ? Thus, _ pathfinder L algorithm like help you plan things athe than waiting until | you_discover__the problern. The_disadvantag is__that a __ bit __ slower _than the _ other algovithras * Algovithm _and_its Basic Con Yo the Flo) denotes the cost, AF choose the _nocle _with the _lowest_{(n) value. a finde _g(n)t hin) gto) shows __the_shortest_path’s value from the starting node t node. nin) chows _the__ heuristic app 0 imation of the value of _ node.ongistency 0 ne - 4 A euristic 7 eco ! iven if the estimate of ae r function __out__to be_equl los than distance __ between _— ai geal (n) and ._a neighbouY and_the cost calculated _to reach __ that neighbour - a Complete - A ith _is__complete._as_long. algorithm as + |—> Branching factor is fini is finite eh => Cot at every action is fixed. imal- Opti at search algorithm is optimal if it follow” __between two _conelitions. itl ac ci i a Admissible:- nee condition Tequite for optionally is that hin) should be sits an admissible heuristic fy pt tye search. An admissible heuyi ti ee. optimistic in __ nature , T Consistency: 4 required Condition ee consistency __for__only pe HPA Comenh| the — heuristic function then A’ Search will ala least cost path. 7 Tne Complenity- Comp jenty of A olgyrithens lime. “| depends on heuristic {unction , ond the number of nodes expanded is eyponentia ty___the depth of — gplution d. So the | time —_eomnplority is (b)*, bis the AN Search pac! comple Kiey _of alooxithn —_is__O(b)4. | Example:- : b 7 A ~adual— (5) —® ( 4 eee eee @ fet 5. oo __-__# ee hie F 3410+4e1F 3y}y6> 16 YrS4i = 20 [spa f SB B2eé yel2ths 22 | i Seopa FE 3474 tue 16 SCDE 2G Br7t2+5+0 = LF _ Optimal and___ shortest path: i $9 C2D2 Fag it must provides the _ optimal a as _it checks _the__ previous . moves 1 as well. | 2_ways:- | forks in two. ways 7 —» Overestimale — _Underestimate os Pnpores- How tp mate At admiserable L oN ! & ties & h¥¢n) under estima t 4 ‘, n) 2 KF (n) vere cf; 7 ‘4 : esti: U . Stimacte L4 Value ac 4| Example ie ui) ae Ho feet Sy = @ cee 7 pa ae nla) tin) 7 Case I: Overestimale:- (_ oo 40 to G ki 50 a) h(A) = BO | sis h(B)=_ 70 w@ ® >if0 f(9) = 200+ B0_= 280 £(B)= 200 #10 = 270 $(Q)= oor5ot O= 250 Case I: Underes timate :- : We may move to other node.=a he nA= 30 1 ttm wee LO _ £(9) = 200+ 30 = 230 m £(B) = 200420 = 220 2 S £(G)- 200+ 50t0* 250 ie q a ee £(G)= 2004 4oto = 240 essai eb iit optimal path. $9024 - Greedy Search A Igovi thm:- || Select ___the _ cheapest path that You think is __cheapest___at first sight. ee Quickly provide. solution but not sure ‘the solution is best or not. jt Io gieedy search expand the node ~ which is close to the goal nodle. i Closeness is __ @stimaled by heutigtic . Current__state is denoted b Estimate the distance bine state and _—_qoal stale nr % the value n(x) then tp lower is closer to the at node Expand the goal node. node. 5$2 D3 E> G & Puzzle using Gueedy Search 2/8 [3 t [213 146 14 3 4 7 5 tee Thitial state Goal stale Solve by using _ heuristic1 \2 1{2)3 8 > 3 4 | 71 {6 [5 146)5 use At we e inthe _puzile game _, the only difference : is that we__add the level ith no. wil heuristic __ value. a 2-9 he Bet gin) +t hin) cot / level no, —___ v heurictig i value ARoblem - Formation ~__Agents have enough information about 2 to, acheive goal State . Water- Jug Problem:- he aes cmaor, 3L and other fo 4 ten how te NL without ___measuvement- oL OL 2 gL OL Oe 2 | > BL al 5 2L ab aL is our goal.