Scribd
Upload
52 views

AIML Module 3 & 4

Uploaded by

Harshit Bhardwaj
Copyright
© © All Rights Reserved
Available Formats
Download as PDF or read online on Scribd
52 views

AIML Module 3 & 4

Uploaded by

Harshit Bhardwaj
Copyright
© © All Rights Reserved
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 97
spe ke Neonat Nefghbos, (kw) ct One of. the almpleat oud wrsatty used Learnt Aeon! Pts ot rs a “potest hi that rial bab Cho ass =the Asiatlastiy 9, tus test cnsrte, Batning A Wr et On pas amd sie Algor tthm a it dors “not mare pssainapi ons KNN & one Of, Ho {Sat phot cos fav doashEtcation when Preno i stead legs Ox no parton krvitedse lop wh Hs, Tiida pl2teloutton + hk ales called a» Lazy Caaath a7 as wits does not Araes He obateg data ta do amp ion. nese. yo minimal a WO taostuling pore & i apite § act. Lock of fate lizecdon Wreauns thot: PRIN) hazpa LL Ha thokneng dole. To be mor exact, alla W 4 ‘ ° : oD Raining date” 1 weohed piling Leatiug plow KNN — booed on feature Siwflant iy How closely out O, We sample peatures -taesble Ou th o.biafiag set detesmenot brow we Claasd fey a avven dat ott. a d wien — : oe os a 5 oS Se ey ~ LJ uirde| dlowd be Sed _ofth oy tp didat st class of squpret oad and lbh SE k=3 tt G nastgned to cecond daw becoure thon ao turd! fh angles A ore Spare awards dhe Wan til - AlgorPthin ee ed _) Brput = “rotating datocet Ty dftamnce moetatc d ,” nor of Wtorset inttabbou, 7 ub in Pyedteted class ov catego Prodthen : fox text Aansple/ instomee “C L_ Fey_ eoch tnstauce £ 10.7 Compute the AStasc0o beturean the teat fuatuce eunsy othin Custowmee [ty the heininy Arotoact: urtng @ apAttonrce: moEr8C » fbb ean Aetonuce)) 4 hetureen tum notds un tho plane fs cord trot oh Cy) a we (ee Fs Soxting tee ditance, th an arconliy Ondo 2 select ine fice kK ncauet tarning deta witrneey & +e tet betane o. i) El 2 Prod tot soe class St beat Pat wa ‘ G He _K selected Ca P OAL votivy i = nth he QL. Consider the seed + trainkog dolrvoat o. rr ahoun fn Tale - meh dosintbes + © peo and iividued cbadents SO 1 ek CGP A cbt red & pate ac “the t ~ penPhuites arc CG PA farsrnmart 4 Pregect- the tangek yantable & * Resudt” which & oa AS cps volued —vasialale Hot re sigs vabstes ‘ Dray! On ‘Eatl’. Bord m ine poniornrouce Ola tudent _, Clean des ushother 2&2 Shident wll gaa Ow dash tr jot —opuiss The teat fuatance given fol, 4o, =) Troitng, Dataset T Dolo} | No. capa Aaooamacrt — Pvos'ect Reosutt Work 4 qe2 as g Paap 2 3 BO ir Pad Bs gl g Fash ae a 4s 5 Paul S 6S 50 4 Fotl 6 8.2 EZ 4 Pane + 5.3 Bes Sj Rail g 2.4 9 | q Vass Sol :- Given a teat fratace (6.1, 40 =) 2 6 Gok Q Cobegortes (Pasa, Feat L] calle 0s Canresa, we Akal to Une re Duaintog set to Aoaaitay Hoe teat Prstenco aah y O Baeltidlaars Avance » The tock & Claceibication ta to anton & | Cas.gronry ow” ase 40 an Orb teary Pokal : Min oY Aseto kK = 3 (titel vabu & 4 | K show be 3, Glaal Max No hywt- Date i i Step 4 i= @lulote +e Eucliloan dotance hethucon He teat Protauce (él Yo Es) ard Cech a. Thaiwing Pustances * Table = Cucl®eloan Afehance Sug: cara] Maouneat] Ryot koa} GoD pe i ee 4 gf ee Z| Peas [x (42-67) + | 95-40) — Al (@-5)2 = 6.2043 2:| @ eon | +l fy (abe c2o-Yoyt_, + (e-s)* —_——+. f r ce = 0-0 tS : | ex) 7 Be BB 3 | 9 Pass 41-012 | 4, 6 as 5 Fail 5.00] ; 5 6.5 5O 4 Fotl 10-055 6. 3 2 $2 y Pass 32-13I14 7 5.8 38 5. Fath 2-022. 3 g.4 q 1 a Paes 5 1.23319 Step 2 i- Sort the dotances ah the order, and colet the fiat 3 nearncat fatnivg dota unitauees to do teat Gite + $ | 5.00! Earl | 5 | 10-05 8 oh = QQ. 0223 Pat Dale I / we take the 3 wcarrot rat gelatin | Hesse we i aa itnstownees _4 Ss avd & Leth Amallead Ad OArees » | ote pd ix Prodret re clare tro teot T ans tance buy ovo Ton! ty Voting : Pnstince 4 Sthe class fay Hat. test peed tat ed “an _* Fadl’. Qa. Clas ipy slat 212 a aatsalal is goed ob vot utes oh, a % gE aac ‘Strout Foun ‘hatalag Aa naples Aw faaen Bs i= XL X2 | Y Actd Dusabiliry Strength | _ Caanifiaation 7 . i + z | Bad i 4 Bod iS 4 Gord ph 4 east the wadertel_ dy be tected ghoul pane the Laled robo wat wWeth X13, Kas F (lz [czas (bans 2. Alb = 4 Stepa:- Aetantrodion of veareot netatrbers based ow kth vatninuum — Aetonce ei HA EC. Dax Rewmk Board — Inckuled 5, FP On Dupree Weigh bore 7 + + 4 4 Yes , ¢ Z 5. 4 NO 3 4y 5 | Yes | & 3.6 of YS ee es ee ce i Dal Step :- Determination db category ¥ of, Fre vicarest ve falb-ours- Xt Xa) -D_,| Rank Snehuded | Cateye Tol 3 Fz) tema? a neigibeded _—_—_ — [ — 7 aa — re 16 3 Yer | Bad + 4 25 q No.1 — 5 4 9 i Yes | Gord 4 ie 13 2. yes | Gerd Stepy i= ating ainuple ma porby 24 the Cote gery ms maanet yaighlons . ine Lowe 2 ‘dered “and Uebod’ 2.” naatenal a claasified’as “good” cole ye: 2e Festus KNN OF stoner all the dataset A Classifier now caacs baad on O Linatlastty nacaaue such as Guclidean Mitownte AU “hakning dats ts trad to clesaify every Aingle dota luatancel. Ae Wetgtted kK-N IN Claontfio i- Pte oF oF St B an extensfan of kK: NN “WN eaatbion, pelowmante te tS eb ly 7 apy dted. mE three feces Birch as nor of _uspiosisaal seelghdeous (fe Sd echione k) distamee Wc Ake wd dectsfon sete The con co pt behind the nlelphited KN & thot kK coceat uctgbbouss te the test Prato aso oratgned ahh welt as Lonspored ta wetqewe rat farther Quay fuow “the Seat [nstomee “The pundple & a pir 7 voetgids ase Aoverstla cvoporHonal ty ols tances, The eclected k neoret notgibors Cour be ani tated form wetotts vith wears thak all Butam ji Cur he BAck Crverwua Hage U distances : } Date tof gicthyry Due Input Trost wry dotosek a, Seu ome WJ fle pe Dictwrte Want ee Fg org sea, oe ee sd — : dS) boohtsad Toat tna tonace. i ho-o, hae ; Outputs Prodi za eae ov cole gory Frafutug " date cot T dy Cooapule the Attar ed ween ate tect Foy cach Snotauee ugh) 4 wk He Saponee 4) utouce OA Cuero en. “AX 4 a duitamee Tact eC. (E-D). U Ex Gost Ye ditanees a ascandiza ode 2 acted de tat ‘kK wont Batty dota Gutamces te the dear aha oaree Zz Prodi the Coen of Hao. seat tastarce oy weg ed vothug ean faye by Cannpute Hee Paverce op e2th dbtuue of ue ‘kK! eelected neorcet As trrrcte. ! Ped wae dunn de thre Averett. Ly Compute tre welyht loo Ae dey Path verre dbdteuce by Ue Le ; Ly Add He wel Uns Aas Cass Giza) by Prod sre claar & Cheearng the Coors unth mon value / vote: ¢ ADL Aastasrrcet — Propect Result ie @ Paah ZO + Roos 3. gi gS Pass i 6 as Ss Foil, Bs 65 50 & Fed &.-2 g.2 $2 4 Paar te 5.3 3e is Foil oF 8.4 a1 q Pasa hiven the post Pustamce ~§ $6 60, @) aa at Cclanaes ( Par, Fath)» uso Hae tbatntvg, datnsst te classiby ue teat Crorounce wating e.Dd ° %& wet olling, funch Mm. Step 4 :- : fastgning K= 2 aud caladariy Cudtean” dur bance. Blea) S No] CGPA| Acaocamen at | pe | Roout | as p! ‘ es 9.2 go _ ae Ne | 8S 0SIST 9. g eo | v ae 20-0289 ates | 31 |g | tes | ab oure e | ¢ as | fail 1g 3S0S St bs 50 & | Fat Jo 2 2636 é. Qo te a = [Q.056S 1 -.2 = foth 29.2764 8. | 2.4 al gq Paws | Bl. 0UDS Step2 - - Gost Hae distances Oy tha aacending _pndiew & aelert ye 3 nrarut trating data ratoances tp 10 teat Pnatante 1 the belocted. Wook Nes glels evi an2 SrA, e-D class 4S LO. $2 636 fail #4 18-380 Feil 6 12056S Pass Steps Prod ver a Choon of He teat Tuthan2 wereld ed techn dpe fon Le seta’ wakad teks [és] Za 2 formpute He Puverte 5 ead AEN ed gl the & Aelocted voorut fustances gs shown Snverse Ovstarnce - Sratamnee e.D ene Chery - IS. 3807 0.06502 Foal, e lo: @ 2 636 0°0923% Coil - 6 12.08 653 0.09244 Pare’ . Bud tee gum O tro pAvercos: Bums 0.06802 + 0:042340 + 0-092 = 0-240 33 2 Compute 0 weal log dived mg each Evouo dotouce by Yo Aum: U Jnstaue| 2 -D ID [wWetyit = Class I-D] sum. a. IS-380S\ | 0-06s0z_ | O'QtOSUST Fal 5S [1092636 | 0042370 | 0.394 34z Foil 6 Tia.0sies 0 08244 | 0.345104 Peat rn AE Measunt Cloortiay Ace ley ee ae ‘a et ! Nrpsre re i ferent Pers onnnanee nact aes —Suipes nts sel en OLgonttemns CLP Clasipication Mace rotates Q Ls. Condulon Matitx L> Ac CUR ey Ls Preas?or Ls Recall or Sensitivity Ls Spe ugpcctiy Ls ' Ea ghee 1) Confusion Matad + Ot yre mandatory to Hud accnay fou ang) parificotion Avidel - Suphons” wer toant te make a prod (et Pon On. & on iyof Lb ube wn. wot bared er some _& Pornoters: So - we ut past gn Aw 5 tang ot Vartables 2 deed buys — Compud er. & 4. does wot bide — toraputev BO. : a bene se Inutle tre niodel , we set_¢ owe < . or & te be a tuples OD Lashed Dupes a Date / I voted thot tat data Aor an achul outped - Sue Conpuuforn Mactan gives tae ef P between hy LX Atctal Vv. ces Achul pre ‘Coad Cra) seed Re tt ba [wil wot bug pork Ss Compu Cy Connped bebval, Yes =i Pe EP +s prep Mehl Wd EN TN > Frttn | L v P+ FN) FP+ TN : Sw thi table, TP dhe achuol & predited Values Que Bue. TW Ane achul predited values sa ore False - Ye Oto wn dae Exrsors OEP RN Z i ( [ = ‘ i f 4 4 4 4 7 “ f a Date foot 1. SencPetvtty s- OF Ps aleo GUed as “Recall” Z all = TP “P+ EN St Bw A pret OC that producer a Tine postive Roebult usthen used on a tert daotacet- at ts aleo kixewn as Thus poartive Aate. } Q Spect¢t by 8 OF produces a Thue Ragas vé Agu On a teat datased St ka Meaae Lotuith & dalse wee Seidl th be pols. = “TN sth ch. FP. Bs Beuwacy i= Foy a docaipion Acuunsett dbcouse Pow connect ly iE funetions = saa TN a rb EN+ FP+ TN Date Acutacy appoaches to 500 "fe uten EMand ! 0 ¥ EP axe leat. 5+ ft alan brown ar Poe thye A Perekaton t= ‘ Prrdtet ye POWwey © OF tel us | had pecpoative ave TP agantet all posittre Kaniples. ie 2 precicton tol ws No- pe patiend cade we dtagueted as howting Cantor, Achill heal COMLLY + U | | Prectclorm 20° TP TP+ FP pre catcfon Cndicatis Aoous good Clasaittery i 'D pacdi Pte wg the +ve Choos- 5 EA 4a of b= Moore 8 Somedinnes uf we do nef unteh to Calculate pacai’cton & Recall separately, we Can Coast ne torr _fvto a cfugle watt - A_Comb’ votton sf hZansente _ Fear: oe, proces Pon & Recall & CGled F- measure Or Fs Abe. [a0] ee eeeeeesS=E" Roe & AVC Cuswes— Dota yt > Roe tsa vPcual means. of Chieckiug the Ccerinn-ep & conmpastsen 04 Clow Pte Roc & Acnattiviog [| TPR) and a Blot & | U dpeution ty CEPR) tow w getven nrecdel = ants Th no” Sc BLED f= otk b Ser 2p be A “eA xan i = Speegely (FPR) j Fav ee le Nore used Foun claaw dion ABs dD Osarhro 2 D 2B deol case Uohire Auc & 4.0 thon srt hral lye Ct Coun age froan 0.9 1 Rost Th ne peupion Ore © cade s as : re = i ‘S| ecw \ Yhe Claastper, pro deen % based Orne! | thaoahold value - A Abrcabotd 2g 0-5 ope Ko probalostihy ty Closes O ow ods 1- the Woarhed— value cot be tuned contol te Artgbsto ow Lower FP 8 EN. YAS OG usepub bin we ous aore cA CARO Fp ow EN Feu an expo feom he otter bept howd Corer St we doure Bue positive Coase s we nowe +P & plot o potut: Set a4 aw jeler positive @ Cash, we Move aiglt & plat - Bes pro Cesta fy Acpeated uA he connplete: cue hers of Rot wae 3 cleaw 40 wine Aeagenes be van dt shours clactfion te be lean acennate- We start Anifally - Leprnting | ua lea rbeg Lehech @ Cupowtsor Bache ov Wours a Ae we wow Supesufsod o uN a 3 : mo chine one lobeled cloth with comet AR UMA Mien Het the tocbute G peortded uth G_new data cok ¥ doth zn Unet Lupermiaed leoanivg alone Cay poduee hp taut outcome fem whe bbelled date. wtth tho halje SY BoLuvua ol Rugiatianlions Qobyts Fadi under Suuposirieed machine Leteirerig Gon | suo < ystenn Pes to pred fice & volue ban aa Cnput hosed en pevtou Paupnodh” on. hg Rema Sail ortont Chorodeisho, = Ae gt n1BO OWA = : |. She areponraed obtatued We musdel Qe adlwrouy’ gyasnbifats ve on notiune. a. the eb Coun be Covstucted only ip pot dita fr gurotlable. On ity aA Stapit'cal preyed Bnnple _ Lvoosu Abed Jot ollouss us to Summarize & shy ho salodionshelp between Two Quant dehve vaualles — Bt] os Dote of 7 4, She doch _vartalole deusted by % fo a praoadltor , O% facep evdlout Vantabel © : pan 6? Valbable ys avaporue OF depenrdont yoxtalale: Aubttomet Ow . a yam | ys artth 5 jz y= ak - } Hs, 7 Z ——— Le Xa 5 x Levee nim = }ncax parts 4 1 2 poy Dytls / um? %-axu “Nom -] fue & — Wnean : nen -\" rele oualep ents TA Pancho Avs G4 L£y4p ‘al 7 4H 2 mie powtet al fren a poe CA Date i of Reg Veasion Methods —— —— = Linu Reareastom won lProow Loge. | U Rug Yee on Regrsssidnr a t —— Shale Muutbiple Yas \Pueon nig vesaton Reg reoare : Bas v { alpen Purpoce Regressten in pe eee To tnd. the Fund onal Arlotionahip betureen HE torget vantaale &- a cubcet | the Aomalantng Aint bul es vontatned vee dlatacet. Si hubew to find Uy Meghlfgt re aund Pudenpret Ye dependency able dorgeb yartable On othr vartabler- i value We target “ le ae aaronal rolhienehip Sdn fied- & pe Valine Le expanetey apansables- “eal usthon © cpa ate wey Qe ern js 4 3 do Wk, We wood tho fro lad in kyuwowen: to ko tho avlatlonahtp hb an otglt and woefatat- 1 Snput_decta Srosaples 4 he te aud vse gect a (arate o Auda enabiife yaodol + >. Fraud sho corpfretonts Lom ano wrodel crsated ( aud estnbligh sire parcctronaaet cal £5 noc on\ trolug thease. "4 ‘ Go Ger @ Susman of tre Aolathonsh! p wurde ‘ _ te _kuroud sire Cuntas ge canons YN prodichin - a Sue aurg exrou Called Rea tducal _ og eg it 5+ Predich wo swore bai N ov peopl —_ ee Oia eaall dataset (0% an op) Loti abate tha Verean Augie? on model bins _predich'ng wet glat 64 Yao poco acsyen de Wank glet T- S0ta 4ample Shoutug bulelt % lsefolt pecple ® U v Wefgtt Wwetalr \s) wo G Bawa Ay 3) 1 de st 1@6 q) 128 at 136 SE t1t4 £L 163 2 Sd 62 Vo) 43 She. po gearion Model wrth ey one prediitor ti Vanioble aud the segnecston “function fi Linear. { fr 2. Bot Bi XT Value 4 Date yo Ne pe a known Cconstank Vesponse Varrable Conrtount : Volue & & Ko Ve a kKwoun prod Peto vaste - a Pe aud pis ae Pararnct bck. To thd RE Qood eatirrat om & saqeenaton parantles, we wisl use, latitod of tena Squbse Po 2 By Ot pabculel so ne Hiab 4 Veen £6 S volue » v — (Pot Bi x ) Qs £ (1: — B. = BAP) = | see ng ® Whe metiod gd, lecut Xe. | nl 2 es usheo X aud Ys ney vapane of X? and Ye ebteass ob'ions nvapect veby T = Dee wie taliudote by amd by we Cow ind Wa _vabue-oh y foy On given %. G fyen the fi fue peraey , he sige fae platen can be predicted. Dbfaka Une aetatimahip between brotglat aud” aychget! S, shuents’ Sol tei ot | tautatah GeO) 1-2) Oe F)G-*) (Conk) Teg 4 iS) | 63 | _). 8 }—-2.3 { E44 | $34 3 LY 19; loo.2 | 15.4 arpiiy [403.08 Pep iow. @ fas | Weqy | 24ee4 3. 1138 sb lea =| Y 7 | a) | 32-2 | 25:74 Q2E.SY | 1036-2 ¢._ 1129 uz | 35-8 | aga [Wey [ees t list | ct |-iz.¢ | -83 iw-y_ | 3168 x. |44 igs | 25.2] to.7 | 2649 64 | 635.04 g._ | 162 Ee q.2 | 6F | 61. 64 184.64 a lis2 [ez | -19 {| ~3.3 6.44 | 3.29 io (1) 14g f-22-8) 17-3 | 394 yy [easy : J [2 [ve > == 153.9 653 REVT. & | 3424.6 ° a ° a be 2h - xX) (-Y/ - oe 4 = g24s UM =F) =n. 6tYS | nee He Wredel & vrtablished we weed te congtom, redethey the viedel fc aeed cough to WAKE. o déchie in future -+ ; Corrslation St & an inp: Atahsheal fernede wat auautifies tre Stiengt, behteen tren vartiaks - nis Va bus. Isles th “deteumintag Aow sir variables. r= aw 2if> 4 i ee CEPTS y= a N — no-@ fusteuces we Lave to Compute the, CLoskpiciont - x Quid 4 ano Yartables - U Compule the correlation coeperctent fo. the Fane dak. Tend aalahon benween Hah g 0 Shutent - S.No. Hefabt Welglt a ca dt (ty GD) . : ' is] wz) asi3 99901 3444 a 44 Q\ Woqy 30246 6561 8 123 Sh 7748 laouy BIDE 4 Ieee oO 14926 3Bus4b& @2¢) eS 12g Yd Gore \E384 ~ 22049 6 136 ot Ws 2 (8496 324 1 14 6 13604 a204) S336 g |k% FD WG i1t36 Sie} elz,) —S-No | He fold [L Wefold | Dt ca ma gy ' (% iw is (G2. dy24 {23104 | 3844 Lo 13) 4@.. [62ee |idtel 2304 — 7) = J = = : 5s 5 \S3.9 65-3 103081| 240¥724, Ys We can evolrote Kant Pearson's Coefprctent te find out ‘f sore & Quchsniy rotation hip between the bakgict aud web kglct Oa persen. Rs.k NS xy = Ze2y = ee =e ro: (40 x 403081) — (1539 653) Cor wor) hay | [Corts eI =, ory ros (0-74) Blea Doe. os 7 Dbaduante @ Usfug Por | \Viesy time Consenntuy Regueres lot ©, rothemad y while progeannnticg ; cal caleutoHous 3 a. pe R- Sqpare St wrW ave Enxhomaads on obout He gordnes ge HE A rauaoe a Linon Acspenaton medel. Ye Lotloutug values are Conrputed for R ’ U t $sqe Zar Bet His ) peo doxtanera_1° a % ype fom Tt bak a am 2 = °o Sum Oh Sqpees aor Foot reno — _ ese = 2 (4p = HF sh 1 SUM Oh squantd Roar dacals faxax} pe a R= = _S$ Ro Cts of 1 osy Ufgler valuc & 2? Odile a stovg Vucan { 4 Relations hip a | [—s For came dotece® celadce tS y SsR, $87, SSE and. Re =. [0-94548| —— Q] du te given Cov le, we love Pye wks Sales data (1 Thousands) » Apply Mcay Rosreeaten te predcct the 3% aud 9 wobth - galef~,. we | Ye Q aS) (Sales is Tho wancy 4 e23 a 1.3 3 2-6 fi 3:2 S 3.8 Hee, we tow & items £21 2,3, 0s! We know ! by = xy) = (x) CH) == Se ba 2 oy) - b © W i [cae xe YI A 1.2 a es 2 \.@ 4 BL 3 Denke gq qe 4 3.2 LG 12.8 S 3-@ Ic V4 Gom=IS | Sum=)2-£ : ss) wom sUuy.y xe Z| Ye Ine 1S | oe x ye S FT od ¢t =3 = OSD y= TE s Q. 8g (| Ec) by & 99 = B(2S2) rere ol + 3” by = 0-66 bo = 2-52 0:66x3 = O.S4 — - fis vis g ae ; Linen = bie sad pode a 2 By. u o = tnlependeat Model Acted owaletp y= bo + by) o 4 =e O-S4+ 0-66 % Pred ichun 4, FY useek sale unl be (%2F) Ws 0-84 + 066 D. yop a fe fy fe] ft it] + |e | Poo a peed | Lyed Lit xnJ Lén_J Ys Xb+ e@ hos Nez matrix Yos X) nnotrix << 2%) Column vector ew Nx) Column vector Example 5.2 Find [tucox Regression efinas hats of treok peodirct doles a St pala - (ORE) [ealar ts cian) \ i Bi. 2 4 4 8 o Date / / Hesse. ine clopendent vaxsfable X is xT = [i 2 3 4) Sndepurdext yortabl & yiei-[ 23-4 ef The dota (Cau he waliten HD mMatrex Ponrr a f Fy Xe [Et] A Beck crtumn bead fen it 2 | setting bias Lf af 0 = - 4 ( vot th i il Bi oe Ci x [vo rae ye TS = Hi 3 Regtesine gahren a az [(xt xy! XT )7 a . 5 Ty). a4 2 2 2),/4 2 | Compute ( X ) ne a { =p Cc — = [4 to | | to 30 | e§ [al Comput & Cxtx), . f-4 iD T LT 30 | Le Os | ~0-5 oe | Compute (Lx vx | [ts os )]xpt 2+ t] C05 O>2- | iz 2 3Y | =a oe o -05) ~p.3 ~o-/ O-} 0.3 | C Ee [ix xX" xT) ¥ C= A. oe a r 4 7a os 0-05 [x] 3 | 0D -0-1 9-1 9 | [: y | L 3 2 “V5 ee ee yr 22H -bS oT e Nabdach ou Reg vearfon retired, 0 oF 4 1 Viean Abcdiils Gney = VRE ; va \ y = rat £20 f b 7 ae Wh wot pee bed ROnApled 2. Mean Sqpaned GDN .- [ALE ok CAAT postive at ceuy t 20 1 = oA _ Ta aan ce i202 3. Root MeaK Sqpose Bxov :- RMSE RMSE = AMSE, = ain og Ne Po RE 4 -—_ Conatden Wo fobtour wg ot wlag Sai fet tow prd fciug soles A Jie Stems + Stoma (22) Sale (yi) yy % Ze 10 33 lod ty l10 Is {20 Conard edt two faeuy thems Te oud Ly usterse » BO aud TS Aeapectively « A Asarpanton “ napded predic We values thems "ic aud gS neep. Clute | Mae, MS, RMSE ‘aud CV “Sot == The test famples axe given Mt Test Items (x9) Ackulvelue | f Predicted Ca) BE re) i é K 34 ko - | es Dao =f Ae Logit 1c Regrecaton _ ny lense Ras nuaston , we predict Hee nunprertcal ALA ponse aa eo at qultable fo prod icing He Categonteal wantobles. (lao fra Lina Pag: Ce COnput os Hae exach mubbut value given yo fnpuile - Response, Vastotale Y -f% Os COntinours oy totive vartaele We neod 1D Petia A&enote vasiable | Whether @ person a mole av female bana On they lotobct- 2. 38 Stidents " walantined on, oak based or eutiow Crantiation markt - Here categwvcel VOetalale oust. admitted ov Not: 3. ‘na Student being pass ow fall based On wanks Secsed. Wr Logetie B Regreaar en & Used a4 & * Bhuowy Cleastfie’ A _nonke by predicting (G| Zoos Date =f 7 | ‘byes: prebaletltty of ele vasetahle- ye dtncady Regacaston, yw prob ale Ly u onedtcted Py eo). Bot Bi PE he—prele.hits buss ase = Long). aa Hh hasge. Decusaae = 00 te 8 —4p Fadl atop Repateruin we bultd He i 104 a PP — bog po _ bib = 3 A+ By ty + fg Siete ait By Lie - We can Actye Hre Logit equation fov_ pe Date I I. Pic ep(dt Bus pie * apex) I+ exp (A+ Pizi+ Pa%2 +t Pe XK) Moove e4pn con be siwplipieal bit dividing NA D by Nurp. ipey fa get: py = Shia funchion iiean fulton fusthien ot & aw Non ltlaan dulletion mi re | ; yor)® Datoe,, ff @ Logo veoaten ertrin falls fy laa» 0 ov Case 2 Gas Slt "Ee System needs to pedi Te a a efaslbers ov not based on (tee ae ep yearly Aout vgs [in Lakhs) gpeoup Customers having yoorly ee tue O ond 6 Lc, Gas — chnsfdlesed Jov analog, i = 7 Tatole hodadet shoustv4 amo Op Sasreeg an '"pon- Ae4 oe fo Comngs (Lakhs Loan (Non- 0-5 © OTS O 1-00" o TAs 5 15 0 “TS 0 eas —f 1 3 ‘ Aop |) 0 bh are ou ° RQ | thaorited cole por ., _ ; a ' te oe ) aS le J : ee : aie , S25 1 ee Ey , S50 0 oo and , Fob CP = > oe | ) URU be shanod , | 4s wn Aifpenunt y 4-S ae } doe Aust = Thoug athe qe. \ g|w 7 7 ae J : aXe Condition = -— 4.07778 = alts Soye. ruse. cota oat Catoud Pato log She 22|.ah'en to gatimate te publ aus bity ey beng a bean ron- dslaulte B ila “ng L Vion = ‘&as| Hhns = ge nea sta Hat eattacat pad iy bleu tila nen — wR Post. “ befus = nen= ef autte T+ €xp(— (T5046. 2 = Ty. OTT} S fowtlanly fov @ Customer unin 4 fakhs —Soactuy pest eat tote eatfasated pee of lonn ——Atpougnrenk uO. SF- Man - default ef lePaslh ey an table - Hose Pited valie, >> 0.6 fe taken as Outil 4 ele eutbut O- ( ( B f ° « ere a. a Ts! | ( 7 ! ‘ ‘ ( Predfeted aud achral value of Auten J sah Savings Non -dqauthev/ Pe ed Pred ichien Clady _Acfouulh er Valine 0.5 oO 0:0344 o Oss ei 0-044977)4 oO 4 ° 0: OFoes ° 1a (e} 0. low ay 6 2.5 0 0- 13% 337 oO Les" ° 0-140 E26 oO bts 4: 0.190624 oO a oO “oO: assbsy oO eas a, O- 333810 oO 5 Oo O42] 602 * oO 6: Z 0:S1471g3. I. | 3 D 0. 60F 3249 \ | BRAT Aa 0:6Q2S58% | 25 20 O- +464s4 } : 4 44 0-9444290 \ | 4. as a 0-910 26246 | | q. S i 0-43 66123 \ 1 Es 4 0-458 60u2 4) COD 6 FO TOIDC«sN(‘#Y’SWSS ae 4 o hg E1400 1 Date (ontus for MatAatx depauuser / -olepoubren Sam Predicted 1 T PagdPcted O Actual D FN Actual 4 Fe TP (2) Nccuraup- TPtIN. = 16 2 OF i) N 25 Precte? on -= TP 2 g 2 0 TpP+ EP lo NPR = "= TT = & = O@ TN+ EN lo Sensibivitt 2 Te = 8 = 08 g Te + EN lo Cumuastzee 8- Dotto / 7 log?t (p) =) —hy Tre log _[_ PC _) = bo + bi x OR =p) To Solve {Ov p(x), we Can take the Envess 2 oh Ye pbove equation - p(x + EXP ( bo + b, %) [+ exp (5 + bx) | ¥ S fanart del phen or ale ees He value WH TD ¢(0-f)- —Biatekng Nao > a meet ple) = 4 T+ exp f=bo= by) oof P(x) 2 ee, eee it exp (- (bot 5) ) x P< the pred Petey’ Vastable by and bi One on ee Caludeted theoug tt =f soblem Sstotemert Gl] Steurdasd thaoaeld vole —y = tk pix) 2 os ¢ ¢ L © Otho nonin ot = the Shulent dataset Mas entrance masks baced en whith Hey will 4¢ Actected w not. Ustug loyahic Yagreaster the values leased pasameters ane ” biped Fy eB Waals 4 x= bo. Compute We Arautlent dass, t See po) 2 by + 5 % = } + 9x60 = 4) We Can dir His value & Siprnsidal Aunchén ts Git pashobil de a : it ep (= 491) = 0: 44 Sj a Frauiming Sid Value Ghatabad 2 OS zs) Sine o.Yyd 0.57 -* Lath mars 60 ee) Date i Je rieeatton Trees 3— Mee weed! when the tonght ass | Peakwo Go pontthows vahsed vasiable: Ie need 40 — noduchim fn varxkrnce utidh user trundasd dovtahten to Chepse best Sphitting ares MHgoaftind 4 ve Sau dard destation 4oew each atts thie uti sagapect 1 tosget attntlute . » 9 (Compute td. Acari ath‘on Hon Hoe una ev dats. ) fin Hae as Geach dictuct Volise. aun alt ssbute - 8 3) fnaapeacte usefigicted Sid. devioh mn fev pace » otk loads - ) ) Auudats acdig sabyital SH. “deus? fa each r \ aitnthube om Sid. dewat'on A each. att xibede- pe) Choose tno atlethute nth Ligaen, 4d -dewiahe ceduth om as Vat beat Split ottaibute F ' (\ tke beat plit atlath ule fe placed ao the Ascot wodle- Lxtth cash aubhee as Qn ouleeme 4 We, tae u Cond PHon othe Acet node oattPlute + teradin J : ~ aie tiata’lg datruk aloo split inte diffessar wurbseks « ° ly A CIDH OA fou the dibscl “Haw Cotwing che Vit te fowolky oHytbut es unhl a beep usde & dertud ono MOAe, Roiw'vg Latances ee ourathable 2 te Sab Sef | Practice Problems bts yy 1 (on stdon tis diloutus tafulig datecet or fo data fnetances hour Y in “ate UV | Ot describes wae pyensdl peAlossrasnce Cp | ivdfvedual students bared en GPA aud No» fh pesiects done. Ye aga VortalJe Arard? which Ga diciete valued Vartoble shat takes 2 Valuer ‘tee w No: TTrafetug Daleset - : : op u : ie § No Gee | tet : Auras a3 5 | Ne 2 8-0 | 4 | vex 3 $2 | i | wo Jisidese-5” | 2 | ‘tea 2 4-5 4 | Yea 32 | | ND # 6-6 | | ND g S4 | | | ND 4 g-4 \ S [ Ven is $0324 4 [Yes ie) is Ahern a tort Pratauce (GPA - te gos No- & prafects - 4) | Use tre hating cot 4a Cl adaifey oo Teak snstounce - (uewie Ke3 a\ _ k-NN Clrosifren v by Wefgited Kenn Cessiftoy J t Sepa t= Calolete He Euclidean dtatruse betwen tne test fuatouece (GPA ~ TS, MOO hegell aad 0 halalug Pasha sree (OE 4) Sno | GPa Nod Ford | fnew Bitance Paget cere: 8-4 -2}t (sf = 47230 o>. q 3.950)) (640121 i. 1 5.49181 3.23107 3.341 et L.ygetos B fe POH [Hs [HO] | - 0.6 Step 2: Sort rhe AStruets wh ther patendsing Qader, % atlect the fire B_Vooret byeheatag, — doth ahodarre teat 45 i Spasteurce J Euclidean Claas . z 6.2 Ya lo ob Ye 4 14s # Yes Step 32 Medict we Claus Wwe teak Sotruce : Lou woes onirty voting « The deas tov Vue peut dratauce 8 predtet oa. Yea", B) Ugtug webglted — K-N 4 Stephic- Some as kK-NN- Step Q- SQme as k-UN° Step 3 -- Predta +e Clas a, Fest ARs HALE + by wetylted vohvy techag pe from t 4 = qe disse elected rieareet taotruces c Dat 4) Conrp whe yt whverse eoth, ART ce ‘ , elgg | Hoee selected Urarcat hearces Snotaaece| Euclidean Suvere Cow Dintruweceé DO ree 2 Ove a Yoo 10 o-£ 166} Yoo 4 LYS} 0-632 | Ye i) Feud dhe drown @ tA vend 7 % Sum = . S+ 1.667 + 00-6725 F33G 4s) Conrputi Hue vnedglet be Aividung each festeae dblonse wy Vee dum G4 shorn Dnatouce | Encltdoan Sinweree usdight = Clos » Qishwe | awe T-Df sum 2 0.2 = 0-68) ‘tes i o£ 1-669 0.222 ies 4 \-4gP 0-672 0-092. | Yes ANY, BAd re webelts 9 aame Clow Noe. oO” Bis HWiew = gay eae 0 e) Pred the clo» ‘oy hecat ag rooteo ps ute WACK MEM VOTE ; Close pdicded on “te” Z G2) A Covid Cane Centre decdex ty develop a “ Coat based eA system be pardick Urledtion & Pecon well teat posth've Ow nagative based ow Sprustons SNo: Feuer Ry Tiredroy Sone Make Headache Lows Cherbes Cee Tat ie ae V— Ye — Yes — 1S — TO —Yu— WY Mey 2—— Ye — No — Yu — nd — no— Yes to SS 96 3 NO— NO— NO — wd NO NO —— NO >— wo —eS= tae NO No PO = NO NO 5 Per pS eS ee Fs Vere Ya ee == jaa See o— TT Ye SS § hr 9 ig on !o__—— No vO NO NO NO NO — NO nO Theat Pan = Sem IORI NO aR NONE Rasuik — ve, = VE) = ve ee, tve, Te WE ‘ ° 7 tve, —ve vd +0 9 better 9 Chen Vest Ve Fo coae <4! value & Chuck prodichm, 86 iH grad Ou Git) Prediel the toot acoull @ the Errlance c = A he Tived nia: $04, SOx throods - Yes, [Povev= Y, Ory Comat Nea, 9 fauthen = 0, Headacte = nb, Sotuton i hear & Taade ov Srreppne 1 2 ow STATIN Dead £-N oO Yes~ Chert Yorn 7 Ko S:No. Nex Yes Yes Yes No wo No No NO Result-e.p | ° Oo. 0 ° 200 '| oOo |} od ) | \ | 4 Oo ° 0 | Ss Oo 6 oO oO 6 oO 0 o oO! T=5 0. 0 ° Br <0 ° 0 0 : 9 0 ° ° 5 io | 1 TT 4 Date mae) Yh efodt 9 boy ads gs dy fiver "s table Hekgit Boys 6S | to we | F2 | Hotes 0 Gils Cet bE po [RI Flt @ Suttable lene Y boat 2+ fev above dak ° = > 4 Nf uss Xt 4) és 63 4225 Yous | Oo Gt 44 od 469 0 1 0 S625 5250 4+@ te 60&Y sé44 Sum '= 239 | Sume27h ING = ava = o ext fs) 2raly 20 esi 19324 b= \ 2 : 2 O25) = S 208.5 ideas 2 te by = (oy) (2) Date os byt O-FuY 4 bo. = [yu 6142 = e u Regesanion Line § bess fet y= oo Fug + IY 617 2% % ov elas poh 4 Ydno teed 4 bu posthive te gaute Voloste Ay gsvcen that pirson bor Motor & 9¢ %, "& Aimflos bby Waws pb 4, Ia teat dy be regodiveu, elect tts, Maloscn frven bt porns dount port rohan Ww be 4E%- Find. Ht pack, 4 A posed beeig UielenP | pee tok, WL toot pratt v_ positive Cebu or i- pl Huypotthwais hos! ~ an Permce Lewtng visoke | = OL Pf Evidence E = ° Test & Posta'vé | Hupothets h = e pauce a MalanZ.”) = 92% P( Cvidense E=' Tear & Negative’ | Lgpothinks h=- ““Dsagn Ahes net bove lueloste } )2 95% P¢ +7 Ee, ‘Tet Gf Postave = [pC bud ence Gz ‘Test & Posttine’) ‘ Sersidnity | od e 2 Test @ C |—Spupeary of Evedeuce E = “Teot Pos ?tve') Dotto = [0.02 # 0-99] + [0.9¢ % (I~ 0-95)] a 0. 0486 pl Hypotree& h =‘ Peron Luu Molo’ | Bvidenw € = | Tet b Pokisre’) = p(Hypotree® h = ‘Perren hang Malan’) we b/ Evtdenee 6 = Teal & Posinre’ | He porrcats h =! Poraen Ira Malarnc’) o/ plevidlewe & = est oe) = oOo 02 * 0.492 0 0686 == 0. 28st eo 2) sino tetuiy dataset T tonebt g Te dite thesia unt pttsPeutea Senr&ian’ = dhe tenga] Bi] VYortabh % Recutk ytte B Caeriped as | nantanment” Pass or Fath Jor a candtdale GHdent — Atnen a tet data to be ( fasesannk9 z! Avboge , Aaatonnrviti = Yes ‘Poyed = NO R Serva a= iced) fanclict Ye Apart 4, Vt student - Tratilry Rotosot | fue. Hast Aa slgannionT hojed Senuxes Resuth A 3 g 5 :6 zi 8 hagar ae RerkeFou Trees td Mt ha predictive moclel for Supemriaed leasnfua pshich tt aia fee booth Cbarificabion he nrodel, Sy i stiiaa to? on tovtafned & wo tatul dotaiet sR te loom 4 a fies , Ovse Hae Concept woedel & built, test clotn cau be easily doarified. Ciena tae katitug dofoset wet a cet } a, Wan’ egeBin tue fb Constructed dinding tne attstbut é a one a 7 ws me ahs farget Deev for He py beak sBstrurces. The best fesdunt selected based on Yo amount — Omeng He — rae ses ann _guontipiod “ a Endrop te urbtch dh Hest sl « le & “te pa AE tai b sh’ 2 : % * ie alien eealeaananagaeaagagaagagrsKg HE Baste Oat Selemeo Techrbyes D1 | Dats Sutepation & Data Thansforunadions 2 Min - Max K-Sepe (Consbdow tho Set dats. Vis oa, 40, Qn 448 feel Min- Maa fade map the saahs ° Min me Set & Ge & Max & 44- New Max = 4: Mappivg B follours i For _€¢ min-max = &¢-$¢ (i-0) 48 q- 8% = oO Min= Max = V= min x [oes nad = ae : max-min new Mes | Foy G0, 0 - $¥_ x (1-0) +0 = 0.33 | qu-&.% ee O66, Fr, Few Tey ie) "ls x Sef 10 to-2 = -/ 100 5 - ad” re “wz “Pe Oe ( Sd sels. ie Led __ zd 200 = eg) “the Trome Set £ St, 40, VY, avy? Cats te | vapped a mais Kenge (0, 0:98,.0°66 J] WF Nis KM om ou Std. devtah om by set. V: E& Uk mean We eet Ve qe Ve Us, 20} BOG Ronen ®S wapakae: 10 Ze Stove - lo— 20 - -10 Saal “F_Suove 4, to = ; 1D ie} Mean = 20 _ Std-devohin 2 1p Z-Suere of 20.2. 20-29 = O= é 6 TO DB to 4 20 = - 2 = x. sue 9 0 = Beate : a Peewee’ Z- score Oy, Han nraasks to, 20,30 (ez) é _ag~ fom Book Date / / iL) X= teat s+ Hoi sips Ak udhs appidaes = nu wo. dada —— “ Cnsaate Be —deniot oa $K- 2 Chi = Square Test rel ey Gendew Regotered nor a Tor Boys a5. so Gitvt se = 0 Total 260 40 ten Leb Mull. Hayate be Hy when nase Tina let H, be -the Oultzawn ata happnttnenle prions Koweshoa ely > diferruce betureen hay agile Fiat we ull ovate aa Cxpected date (ey iF Ex peeted Datw Gendev Regirteud Not Renisterad Tetet Boys Sox 60 BO xw Tso (co oD = S86 2 Qhts 5oxéo =3d So AYO 60. [op Paes Tor) 60 YD A i. yee Ee) tz! eT. SE ae oe ee 20 20 30 Es 25 — 20 20 = 4.164 Pov dogce oh freacom iss No oa, gategese * : af st pavalae fet +42 Stetishc Po O.0#12 — O-o412 4 0: 0S” wala Resut Fs ES qr fined : Pov eg -) Whena com it f&ppeds heath av! Tall axe the twp outtemes. Heme 1h erties a foureu when compared to tolling a dice ustaich han gt Six outtrome- Hemee Hteapastetion fi a Highey tre Endvopy —> Highew the Unentointiy Lower Hu cxbopy 5 | Lowen the Uneertnindty- tov eg. :- of Hare are 10 dota tustauces , out J which 6 belowge +0 pasitire laos & 4 belongs negative lees then clit a a caludotecl as:— meer 2 = ee = eee Rompe. a Cdsepe [Oi A S¢ doticot has Pusdguces thot ae bLomogensus Hatin _Exdvo py i zexo . of dataset hae Soe thot one ogellg divided “(, -s09,) é 4. Thus Eotort valve rawges hetwery 0 « 4. based on parrdommom oh the samples 1h dhe dataset «tho Stopping witeta & ased en He Exhepy yvolue. | Lins) fF Let P be profoabl tony AlPstib ution 0) acta! ! Sanaples pom _ 4+ BN 66, (pe Pry Pa rest Pr: er fg the Pupoumast on WadAUe Of partabiliny dAGhe busion Bako aSpaie) 2 Cuinepy . Ingo (A---+ Pa) = = (Pi tea a(A + Blog, (fe) + xt Pay bags (Pa) ) Py = Ay chain esas eae bolenging $0 Uae + Tore wo Op dats factwces Or tradetog Lxdvopy dlc (o,4) = _ [% sea) Tags fa Bay malt = +—\_] so ante [Zia] DectsPon Toes Types Pteasena) Mio Sie Dy =R3 C4+5 CART ‘Quest GurEpe x} $ “They lonatdey only -. is _One. feoduve Jatt bute sl spite ot cock at dee clecfston node sity I. Compute, Ertan Into, fov Complete hafal clata cet based aw tro fanger oti bute. : “g. Compute Bigtvepds 2nfo ound Snpervwasion —Sain ov Cosh the attatbute tr Batntag dotaadh aa Unoose MWe ottrbuter fou ustutth. Extvery is Mentmum & Gein th max 2s the boat cpltt atirigute Ye the best split athribufe 3 placed as hs hoot node . 5: the -noot node % pnouded into subtrm with each Subtree as an outteme Q. the best tandttion 9, the cosh — attribute. Koduiug do Gl] uw alo =i ae Bi T, fi j 3 Dato / ey on { =. 2; Cp loge FF t=! Extvopy 9) ere atftbute cdenoted as fotopy = Tnpu (, h) fs given by Entropy — In TA] SIMs Bae - age A) = Soe ES Information Gahr & a mnitate Tak Maar & drpuva wre lnfdamadion a gain ed broach in Paapuni ty nan Orb Tharp Subset 4 dat - Srpormastton — Gain (A) = Evdnopy — Info (T) — i Entropy - Into (TA) Lian) ai) faotan a Student perfomance estes 1 yh Course @ shady & predict uthetho a Ltrdert will git a seh okpev ov wot in OE Hual yaar QD Course. the Raining dodisot T conaits a lo data fuatarces unt ‘olbbutes c Tab. he Yorget dane & “Job often’. Werign a deuvion Bee Jo se. - = 2 S:No._CGPh Sutew Practreod — Tobofer - 224 ben Vey god Gow} y 2 oe tala Gord Medorok y = — Me nrg Poor N 4 =e Ave Govd LN pe ee Yes G 4 y a ae G M ms + 29 Yes G P e 24 No VG G y 5 223 Yes q G y Le zB Yes Pr Gg y lain) lube 1— Solution i boo yf 4-- Giluwlte Heo. Ext Srep = ab alge tropy 4) We tangst Ewinopy — Info/ Target Attribute = Tob offer) = aa oe 3) = =F + to t+ “Dm top 2 = {-0-2594 t (0-08) O- Stor. Tyesahion 4 = Sete 2. i Caluulake toe Entropy Sapo. Gout Linfonwati en Gar \ for each 9, abatbute 6 Tainivg dataset Table Shout tua nod, dota thrtewces cloasifiod wath Tob offer, as Yes ov No fo te atinfbute CGPA tt Evseopy Supornvaiten for COPA pee a CGPA Totes ates] Sette 0| toted | / | Cuctvepey 24 4 | Zege a O 4 | O <2 o z Zz 3 —Eatrepy Suk (TT. cq pa) = 4 3 log BT — leq, } | foro 4 kj Ty + SL a see 0. 2 Sg ae tS = 2 oy 2 2 eS] To | = =a CO-3nT + 0444F) +OFO < O . 3243. Gon CURE = —_ BEOF — 0.32493 = 0O.SSs&Y. ¥ Tolle thous due dota Burruces clasiaiprad picid “Te offer aa Yea ov NO for atintaute Duterachvenuas Sito Sufonnodions fou Suteracdivences Sutero chuencers Sef offen Thopter= 1p Total Yes Ss 1 Z NO a > i; Evdvopy . Sufo (T Suteraciverres) = 6 fo bac - 5 log + | 4 Ih. oO Ee eS S = [ 2, 9, % = 4 gs a | ae SU RL AN OTF, : O-s8Ie + O-B9IF = 0 F240 | Gouin ( ueradivenens) = 0: 680+ — 6. FEqE 0..0atl 18 iE { [ ( “Table Slows the pirnrbey of data. (weber, | Clacsipied twfth Sls offev as Yet or NO for tue atta bute Practepul Knowledge.’ Pra chiral knouiedge Tobolfas Yer Tobegyars NO Fetes « New Gere} 2 oO 2 Prbcaa 1 2 3 Good 4 t : fudvepy _ Qube (Tr Pra chook kuoutedge ) = 2 f=2 bog, 2 - Oleg, 0] + TO i > J 3 ee eS, STF 7 LL = tog. = ical = toy =] = = (oye == (0-5280 + 0-3849) + id “ —& (0.25974 0-46N) O+ 0-2453+ 0.3608 = oO. 6361 636 Ganka op eel Miteridesl 0-0} — 0.636) = 0.2446 | en) O / ty, se Tale Aros Hiro date radanrcer che. Used usth Teoh offer as" Yes" ev “No” fow allafbutre Communitoh'en Skéells. Comm SKE LL Tebopj=Yer — Ieboffe.= no “oral. Gord g 1 e Modesats a 0 3 Pook oO 2 2 Euteopy . Suto (T Conamurntoahion Silt) = i = T ag Ley, 2 = eo E+ 3 (2 yg S ~ 0 Ing %) a + Nol 2°27 7. Z| ut 0. 3604 - Costn ( Commun teak im SATU) u O-§€13— 0.3609 0.5203. o slg w / t By - ° Date ¢ Gat (Coleuteted fow all the otttihut Ee o Attath uber Gain CGPA o-.ss6Y Sutera ctivences 0:-049]]) Prackcal knousrdee 0.2246 Coram” £ kids O.s5253 Step® — Fern “Ha Talale chorose he _athibde J, fom sites Ewrvepy S _pafininwuns fe ook ts Wa xtnum as the bout Split atlorle ute te Aibute Uu CG PA twee P an max. gaty . We choose -CGPh an Root vode - Unico el Cape UWA auttemes Ze, Exdrepi vole G O fxn 28 awd 2? wit all Lustauces Cloarified as SHwet values 2d Be Sob oe Yes" doy 2B Job opper =! nd fer 2%. Heuer boty zGPA

You might also like