Scribd
Upload
3 views12 pages

Approximation Algorithms NOTES

CS3401 APPROXIMATION ALGORITHM

Uploaded by

saivivekphd
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
3 views12 pages

Approximation Algorithms NOTES

CS3401 APPROXIMATION ALGORITHM

Uploaded by

saivivekphd
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
\ pero ieaon plgoritions for NP ‘Hod. Treble © ee ee = San compntional Conslacty funy, tron axe teak tures OF pasbbers. > some pooriems or gecisvn problems fos phidh orgy 1S Yes OF ab: phimi-yalion problems . dr problems ond. mong offes O' OF others Que Seon rial tine 7S Qentially gested by 0M > “rw solvability of protloms vn paby no alarninis KC qwing Moduies - pene the Complexity class oF puddles yhak on prinsially hades CONDE Solved by A non —dalaanranis KC qquaing modu Yon Yose Yok proud penblems. pobyromicl HO gre cold NP- gkorol problem e- ComgLale MTeE gorision versiod OF o Combi ss paved to be n LD Of miaghio™ veses 1S NQ_hod + Eai- (onside & quoblern of pomisnian Cycle - MEME yi nion ode wih tex too K Yun pais as dR? Paowber and Y — Buk, shoctest Horobtoriar Cyl@ 1S o ne-howd problem Because 1 1s Not easy to duane sf the Lye obroed iS chose An digosttyn ahal “KWAN 1120" optimal co\uhioN* in gdavis gone 1S Frown aS pggcoxieation dyer Pisclen — Knagsark, TSP, \ — phoblem os hord os Nereis Polunvmial. dime clgorithms yo Solve these Gip) is Sill ~ echouskve serdy Sparoximnation Ahad ou hk eS opheringadion prblams . wo find approxima sobutions 0 Set ore fe asvocialad wih np-hod pootlems bemue UIs wwry BRKME go wh on ettident polyramioh ‘ polynomich june Cxock ofyor tun SAwin Ng-howd prodloms ayer e Xv Aprooximatioy glgonthms of biel 09 heuristics (5 ‘ PUSHES €. pro(eedsng Soon by pick and KOF enwihod 5" Ct Raasa —> a hmuristics con be defined oS a (ollediion of cone ules daNAwd Fromm exewndn (2 Cont ACbuXaty vohio > se is Buoys rnestory to Know the actraly of wre agua pphiodl solo === apptosimation to Adlutaty YO is dasivd 9S : £(so | Sa > opmoxirnae a tion ees) = iets) £(s*) Elsa) > ydur of cbjede fundion fur SolvHon £ts) > yyolur of objedive fundies for sapvion given 4d ASDF 4- wher ¥(Se) wavs clove appexicnals colulion: ei >i |he ‘i mapeaijakvo oe fC) Sone al Bost pduowich Hime agptorimasion dlyrattwo 7E te achtaly TNO of jhe (Sa) => ALbsraHy ai given vg aprtxinadion Bysottn Cxork ogee thr jo 4. thon aid 10 be C for omy instant oF The afre vest (12: the Seale) vole. of C Ms for alt the Greapolts ((-e a) (+ sro 3 =1¥ Q—c+ A—b—-& = 3yx1t3rI-E at b- A-O = BAB b= y BOM je.( Me Sa 1S 257 cS Loogor than oplimak tour St) Sno) — (s a Bo - pe) = Cae wis a-a) F(s") : ot dy (ad) fron b te on qabrikonly Longe water) te. this dasithin js suiple 9 we bu hos asowhark- The mojor drqwback of this F some os oth which hos to rroueyed tO oe A-a) « mint « fe Rodger valux of Wi, (fsa) sM Fad sa the pphanal vill fal to oblois oer a ee a ®, ss nee (BCs tos) > xe peice Oe eT Ot CRG) eH oz Are-B-D-A= UxBrhHT= 18 pec d-B- A Hyer EDAD A-D-C-@—-A = e273 42> 5 AH-D-B-C-A = AFUT3+H = IE Yoo= 40 _ HE = bb \ FRO a by. Qoaye thao st D-A Wa ors path -° Gaye thw! (w> 0-4) ae E iRx- =) Dud oppmximation algovitin wid) fail to obtaxn the optindl Solus oo 3 pes tne pore Misnbovt dboittun yo the Uutone defined by the fos croix who. sak te dpithen ab te Fidk cihy - astuming sho cites ot Aumba Fore 4}S. 4 B Wom s 4 Ss 5 ie ye Tetra 1b O0) o & Bene Ops. <)t ayy - on") g le Talenor = we 4 ego ree + Ib Qa o \ra-5-3-4a = 2-8-4 3-1 = 66 | 1-y-5-2-3-1 = SR Bech-Q-5-3-1= 4S oe r ® —hascishe_ abapithen eda of That weights - anwdby & 5 of tou Sigh Gon te aes Wncctasing edges tobe Coastvo ches es tp the tnstons ind solved yo tha. Omagh SH Shap 2° owt aris SBP tines, whats A is ee (w. of GF ada he nek CRE on the sovted ody list VF Yis nok (ranks volt of Aopen 3 OF % cde oF Length ess than 0 = pRwwik, ati He oye Gps: pain tu oor of tol qe w astondiny owdet soit Yeo _ {e» G9, &), @O (as)) soul (a:4) @Y GOO ad tep> FC9 eee ais ie oe : 36. Lange F(s*) than orfisnad four s* . © Twil@e ahound—the tre dhovith ' atid” Condteuks 0 minimwn spanning Hee oF WME qrogh Guesponding to o 91V0N wakene of he doveling calasimnan “Psoblerd : Sap 2 stealing ak an onbatiany vookiat, pesfinem o wad qhound the HST sopeding a the wis posted bY- (DFS tsowasal) one * we Be ee te ets Wet quam abe 2 on dint fiom i ob He? occutenes of Ye some wht arigat stolind ona ok the ond oF et =the vlies raining on Yee List wil for o Homibtonio Ci ys We ovteul op the algorithm Beinn te HST pF Ye given gray he HST Ff this grogh is made UP Of edges Gy) 69 ba, GO Rend hoe a DES WadK atornd the hee woK tho stolS ond ends ok O- bd, @, dbo ee icy, Ps eo Eliminclad 2% b, and a ond s (oe ib yids Homblnsion Gxt) th 39 opsinal @ > te Jwie- arround Aha Hee algerithm 4 ronal - @ Qn enako” ak; When For 199 salesman probly with Euclidean cca oe os age Ve md te shew Yhak, for any Euclidean instore of ie TSP, the Yoh of 0 Your Sa obloied by te wie-cwunl- the Mee is akmost toa aha Wend of HO oyHimnal tout sé he Ro < aF(s > Pamowing any edat fron S* gets a Strmmuis tee woishh colt whi ‘, rust Se gyealt oy Cyd to weisht OF the grogks 157 w(r? €(st) > abn Sae-t Fee oydlitis sights tat the Vonsth of the walk obhuned ms) >out)= ™, atl aul \ slop> oF Re olyerthey « Pg ot te wale ted) ue lth oF te tans arts) > F&> h x Tr is the aPplorimakion cle CL elit prrformane vabiO fox the proviem DeES- Derth Ful seo ~ oes awe, than viet of a gag that Visits Cry ody A otis with on evn rumba of alts wth on odd numla of adbes WF & qmgh thok Visits Cour volee : aotth ‘3 praant abgosittu that a a a ; My Minium Y . Spasmung Wee = es. Conneds ll vakos with minimum Possible edge vsghl Fe oon | Xn chock, 0 Problem that fakes on jyne to Solve J Mah’. eve, velar hos exactly one cage: sane bly > Po W. iangle newly: tha shoslst d christofides algorithin is XO fos Yhw Euclidean jroudling se isc. SHoighE juno. yho axpproximno}0% agontess wih o oles 0 -peoblema chaisto Fides Alar’ . ‘pain persowen0sy gus.) Fad a Minimum sponring TR Ai g. cred Hw 08d dpyur ond ever doyioe of veskilos wt 3. pad te ays of & pms - WSN tmokduns OF AL odd dager oot nd dys Oe a ts Ge ceed ~) yr) | => ditt a eae @,O (he) =) S4='t * urna (Bie) Chic) = 12th =18 Eunion tuk = Pet ee A pamiborign Yur =) a-b-C-e- oes @ @& aha 3 Copkivek : tour) j a 7 or Eudidson ‘Pishanes {nskantes, Qood apptoximakons +0 optimal Youss Can be. raghovernent yosttans, NE cabled Lace). Search pynisties donb: ua -noonos ete) 4 Algerie, ‘) Stonk wi Some aluad. tour Cby uss nwighbs yon cash aston the Aaorithen Cxplowes paighbothood nun see Gunent Your be eelauns co. fw edses phe Current UE by pita odes» jorter tour, the algorithen qnakes Jbor7o8 rh He > ap te ne produ OS : a corks bs exposing WS ci) Paes inenen, Alopaithin fos Knapsack problem () » Crier “ny” dems of Known uvights W, wa, .- wn and velas Vi,Va, «+ Vo- noble Subset OF tho and a Kpapsark OF unight cnpatity vi, phd Me most vd Yams that Als into The nopsinck 4 Distal Knapsndk problem => Seve J Conkirwous Knagsack ‘paeblen =) padiond yaks am entiady or donk selec ob a. tom caf\ be aloud 40 glow Gredy algositir fos we Ajstade Knap sac _ Oo As Compulle, Yhe aus Yo vxishk yah > 2+ sos\ ko tems Apsiondaiy owdat of Nilux 3+ Ragas Ane. Folowie® ppetodion uri ao sen “8 Week wate Sorted Vist = Gif he Covent “am on te Vey GAS ville he Knwgsadk, plate a yo the. Mak Vem who Knagends and pso6ve! pthauwine, ust pooaed ty tre 12 yh ie Viko; for ol deen pebt vas ot 2S tes 65 409 for Jam , bal ES of Ynapsoak Ppob! ry " (mds, Ayottor fer te erik rusous Krag Sock problem ———— a . 1 Compl ha. vabo 40 unis Spo valu: ich 2-- 9 ah dams 2: Sot te LS Agstonding evlst of wai0 Vi fwi 3: Rageak tha. fovewing opeiaiion yaltl the kncagenc® js FWA fro ys FAL Cogatity pe no dam is lett wie ceeted \ish my it fra Casters “tase on Nee. ist fis vil he. Koagsock ik onal, gave. ond poowed 40 He cast Tam Be trensise, ‘oie 3s log fwckon v0 FP de. Knope to J4 ful qopnils ond Sop Rn a re ahr ot bs fs ae eo. jp to. regen fon 40. Qin vast of Fnoggaxt » on x — Tr Knapsouk piellem, thaw east Goer ook cllowt us to ot arpoxmate® OLN Jods Bei, Kos one Fe ie 10°F Kis on ie Oe N- ® > The Fast apptox mation —schome w0s Suapestad by 5, Sahel in 1915 > st germrales all subsets of “eens ov Los and fos each one that Ais til ho knapsack adds the yemaining femS as the Gordy gosttn 12 TFéasing odor Of Vi fy; sadio- => Ro cubek of tho highest vale pbtaied inthyy mnottod is velveed Os tte allyathms ole. i a manioeeat lel | Vile: Ey, Apreoximation sdwme K>* eal es ln | Vil Por lege | Crpoily we !0 bee alees |. 2 peree es bee as SIE BE cppersqocton Swe. K=2 means, i Oosder subsds ot fess than 0 (ae e a i 7-2 Subset wth 0,4 OF > doms é ahis fying “the copa Consreounks O# ton be Considaud ond @mdieiy tems si Waa) acoordnjly. Noluos | “yar30 +yaTb | 42 $303 = 42430 | aye | hI444=53 | Seyery = sb | Qt4Qt 30-7 { Wel) Not [Link]| me | ale WI + l | ae: _ No} feoniie

You might also like