Machine Leawning — BCS602 ea St at eo Ph gga ee instantesa 4. Comsides the training dataset of 4 in of Enjoy spol: Apply the Find S-Algositivmn . Enis. Example Sky [AtrTent | thimidity| iaind | ates Faent Set Lo Sinny | Wa any Nosmal Strong Warm) Same | Ves Sunny |. Warm High | 2 Strong Wasm| Same | Yes fe Rainy | Cold High 7 \ 4+ strong |wawm| Some | ie Sunny | wasn High Strong Jory [Same ly. Solution + Step + Stet with the most Speci fie hy pethesis: TnitiaL hypothesis h= Le, bbs, 4,6] ») Stee Update -h wring “positive exam ples only. Nosmal,Skrong Warm, Somer] ™ Same) Tesation 2. (positive): [Sunny »Waam > tigh, Staong, Waam, Compate’ with b { Same) Sky: Sunny = Sunny >" keep Sunny | exalion 4 (positive) : [Sunny shann 5 | Update hh: he [Sunny Warm, Nonmal, Stang, Was AisTemp: Waams Wom —> keep, Warm temidity ; Noamal + High > Set to 7 Wind : Strong = Strong > keep Stacng. Water: Waam = Was —> keep Wasm Forecast + Same= Same —> keep Same Update bh: b=[Sunny >Wasm, 2, Strong , Waam,Same ]Tenation 3 (Negative) —> Tanore Iteration 4 Cpesitive) Sumy, warm, High, Stamg, Warm. Same } Compare put be All attzibutey match on alseady 2. So, no change Finab Hy pethewis 7 h= [Sunny. Waam, ?, Staong, Waam, Same | ae 6. Crsider the Shudent performance training dotasek of % Instances givew in the table, clarify whith, a Student will pass or fail yy the course bared ow b-NN method. Pind (K=2). Given a tert instance (2.1, 40,5) B dad Capa | Assessment | Pagjeck Submited] Result L a2 35 S Pare 2 @ se 7 7% Poss eS a5 ca 8 Pass 4 6 ae 5 Fail 5) 65 50 4 Fat 6 a2 72 7 Fons 7 5-8 ag 5. Foil 8 8.4 21 4 Past. Sstudion. Stept: Calulate the Euclidean distance bekween the test “instance (6.1.40,5) and eady of the training Inte tancer - Foam for Guclidean, distance - ist = : 7 a \ Oy ny” + Gar y+ Cea- Yr ere 1 %2,Xe_ = Given imstance Values (6.1, he, 5) Mr Yards = Values of each ani ning inpteneetemt | PRO) | Rewbl| [Link] | COPA. ‘Assessment Submited | Cs-8)* y Vieqs.76 * 41.1746) ! the fe eS (40-454 (s-s)* -fo-o14as+o | J= (25.0) > 5.00} —_—_—_— at to |e Nero . (4o-50)*+ G4 > flol-te. = 10,05781 TE fata 1] peed Poo | [G18 2)? (40-724 Ta ¢s-7)~ = 32.1314 t I 5 [Pal | (er-sei Gear, Cs-sp- 0.044440 - V4.69 = 2.0223 23 Pass (e Ee a (4o-an¢ + (5-4) "784 +2601+ 16 = $2624.84 = 5I-23319 Step 2: Sort the distances iy Ascend. other amd Seleck the (vst 3 heart Taaining dale ing bines wo dhe tet ‘mstances. Swnee (k=3) 1y 2.0223 8> 51.23314 ' i 27 5.00] The fist 3 nearest training deka. 2 10.0576 instines one Ay o2.1ony ly &ocag, — Fail a> 40.0450) Q> 5.00, — Fal Gr -AL. 17961 T> 45-2063 3> 10-057¢~ Fach SeStep at Predict the, less ofthe. Teak: ing Knee a . majority voting. The claws for the -test- ae is predided as “pi Ff Crsider am example uchere the five nvcks sales data (‘in Thousands) is aevem an Shown below in Table. Azply Uneas Regression, technique -to predict Ake 7 amd at moth Sales. vi (Sales iw Thousands) 12 1.3 1246, 32 33 Po ee Selution - To preset: the Th and al weeks sales using ieee Regression + we fiat fit a straight ine whe, X= 5 Y= sales” ya athe, eCq.0 age Inkeacept de= Slo reStep 1 Constauck lhe Computation ble and 4 Pind recemany values . xj Yi | Com Xray | p o4 V2 ! L2 Baise Be ce 4 3-6 i 3 : 2-6 ; 4 7.3 4 3.2 \6 12.8 5s 3.8 25 ne Sum = 15 Sum=12.6 | Sum= 55 Sum = 44.4. Average (i) | Average (Yi) | Average (xi) | Average (%i#y)) . Lee Hea’ Covrputs ste and ieee wring foxmuta : : : “= OV) - OG) 2 -()* a =(Y)-ay RX ; | , i a= 9.8% —B(as2) _ ¢.es—756 _ 1.32 20.66 ; yo w=-4 a (row, %o= 256—0.66*3 = 3856-|.48 = 0.5H : a= O-D& ‘aeSp SB. Model the selationship wing 8410” Ya eo + 1% ard a, 2 a) iw above ewan Y= OGY +0.66% Step. prardick for ai and ai useck. th Fox usec > Y= 0.54 4 0.66 (4) = 0.54 44-62 (xt) i Gq = 316) in : Fox 4” use Y= O54 10-66 (D = 0.54 45-44 ae | (x= ee | Final Aner + op ywecck gales a 5.16 thousand a uscok Sales 2 6.48 thousand ver date provided. and Apply @ 9. Conyider Dhe 9 Palynrmial Regression ‘and sdve it (Use oxder=2)Sdudjen : : pot | Step ts Constauik the Computaion table & recess 7 Eo Vales 2 3 Hl ti [| i Pye x t ! ; \ 1 Il 1 & Ho fos®e Tek le eo] te 3 q at 4 3\ at 8| 4 is | Go lo Bho | by 956 Santo | Bayi 84| Say Ae | 20 | 2-28 Sed loo |e ! | Hepa. Th can be nebed that N=4, Dyi=295 Sui = 96, Zxhy, = 338. When order i¢ & the matyie | Is owen ax » N SH, Say Xo yj Sy So SP | 1 | Sey Sx se sxe |(| | Sah yo hy is of the form XasB. Oe gmao-- Tk can be Witton ad azx<'p 4 10 30 Xo 24 1o 30 loo ay = 1 46 . 28 30 \00 354 a, 3ao bo} Fle So ] 24 x ay: lo 30 loo 46 ai ag be 354. 338 —6o,75 6.95 o.75 ep 3: “The Regression Squotion is Gjome: by. Ye oT Aye Foe Yr -OFs + ons tone |lo. Covet gure ov Regression tree wring the fAlenting ‘ dable uehicd commists of 10 cata Instances and + = & athibutes ‘Assessment 1 Assignment and “Proyere. The —tangek attribute is he "Result wshich is comtinues atbibute. ashe | Assessment | Assignment: | Pasjeck | Result (4) | | 1 Good Yes Yer 95 fee. Aveaage Yes No To 3 Good No Yes 75 4 Poor No No . A 5 Good Yes Yes a3 6 Average No Yes £0 7 Good No No 75 8 Poox Yes Yrs » 6S 4 Average No No 6s lo : Good Yes es 84 Solution : | Step : Compute Standard deviation fer eady atribute wat | +o he tang alribule +. - ve - os ‘ A we = 95470475445 498480475 + 65+5e+494 \o = 750 to Average = 75 Strndawd deviation = [ 5-75)" Go-7sy + sey + C4575 4 @s-75)+ (S0-75)+ C75-75)>4 (65-76) Ge-75 y+ C84-75)>| Assessment = Good Ss danesement Aswignnsa | Result (7) 1 Good Vox 35 ne Good No 75 8 Good Yes 98 log Good i No - 75 lo Good Yes 39 | Average = (854754964 75489) {ee 5 = 432 S Average = 86.4 (95-8e-4)"+ (75-86.4)*+ (e-s64f + (75 -86.49 4+ (63-86.4)> S Standard Deviation = Standond Qeviakion = 10.4 Aswesement 2 Aye vage SLNo | A seeeement- _ Assignment | Result Gy 2 Ave wage Yea Jo é Average No so q Prvesage No 58 | Average = Fo+to455 i soo A 3 Pverone = 206 Average = 169.3Sondesd Deviation = Go-64.3)*} (¢0-€9.3)% 6s-6qay See Ne Assessment= Poow Li Ne Asecsement ( Pei apo ment | Result () ls 4 Poor No | As Weg bp Pere Yes ' | 65 = illo z RN Fi / 5 Standovad Beviation = 44 oT Standond deviation amd Sota mstances =55 No} I 1g Weigtted Standond deviation foc AS 685 men — = (10.4 +3 or) +2 > (l0.4) Se We) = 4) = IS hte deviating deduction for Assessment = l6.55-53 = 4.07Awe ment: = Yes Assessment | Asoigwment | Reult- Gro oy, yes os Drevage yer 70 oy ool es. 98 Pony yen os Good ye 20 Ave sage = 99+70 493465489 = 417 C2.4 Eee S 5 Standard @oviatiog = (s— 32.44 (o-8 34) + (Cag -33." + Ces- 83.04 9-83.97 ee Statand Quintin = AS Assiqnment = No [Link] \ 458 extent A Signet eae a Geod No 7 4 Poo io ie 6 Avevage No Be uw Good No = ' Prema No 5g Avenage = 754454 20+ 75458 = G66 i) Stomdand Boviaking = ee 66.0)*+ (45-66.6)"4 (G0-Ge.0)" + (75~66.6)* 4 (58- 664) /s = AATze nabances fer asi but Aswigmanet- Standard Mevialin | Oalar Lastrnees Standard Meviatiny and Oat Agere ssw ente ae a8 5 wyes No ' 14.7 S A | Weigttled Standovd deviation Pov Asslanment = 5 (8) +S. (4.7) \0 ry ‘ : = 14.84 | Gkandand Deviating seduction -for Assignment = ; | dee 16.55 — Ih. 84. = 7 | Slandaad deviatim Reduction fy Each Mribute Atsibutes Skandavd deviation Reduction Assessment A497 Assignment Stk Since Assesement’ has the marimum Standard deviation Acductinn LE is cheoter ag the best Spretingy atribute.Const mrt Repression Trace Assessment No | 75 No |75 Ei sob on Jel Jounin Ghee! ps [go] | “se & [Pos 4 a No |5¢ 113. Access a ekideadls peafoamance using Naive \ Bayes algorithm uaith the dohaset provided im | he tables Pardist mehither a Stident gels a job offer om notin his Final ngiar of the Course. \ [Link] | capa \ Tick enactivences | Tob Offer i t nal ~ Yes Ujes f a Bee. ‘ No Yor OS 24. No F No i 4 £4 No No | 5 we Yee. ‘fet pe z4 Yes ‘Yes | ot <3 Yess © No as xA No qe ae 2s Yea es lo 2 ~ Mes | Yes Step Compute. the prise Probab tity for "T5b Offer’ Poor patbaliddy for ‘Teh Offers = ayes = Yo . Patos pasbali'vity for “Teo offr="No = 2 a : 10 To offen Clauws | Ne of instances | Paobabi ae ee [Ne | Ves i Shs 7 ; PCHbO Fe =yen)a7,| No) j , ro PGrborter- ny) =¥,Saree Compute Faequomy mooi andi be keliberd Pasbabitity Tov cath spthe fanlre, 20) Featiue— cer 4 The Of fev=No 2a ie re 4 ° <8 ° ae Teta 7 Aeon? (2b) Feline — Interact vernsys Ukelihecd probability OY Caer CeyPee | F Gr ofa = yes) | PCS effer= Ne) 24 Pearary4 ITeb 0 Ff yes) Ya ¢ = 24 we 444 La <8 oF Ba, a Faeg uemy matrioe of Inteoackivencis z Diberact veness Tob Offer = Yes Tob Offer= No 5 4 g &Lileabé hood probability, x Dita veness 1 | Teter ackivencns | P(Tsb offer = Yer) [rome ofl =) ad Ys “e 4 4s | 322. Use Baues theo equetinn—te © aS || dhe Parbol lity Fy ote hugpetheses. DS Offer = yes | Test data = P (CaP 4 [Tc Offer = Yes). P (Deternctiveness =Yos| | Tsb offer = ye) +P (Dt offer = yes) | | 14 ie | P (Sb offer = No| Test data) = | P (Capa >4 | seb Offer=No) - P (Drteractivencss yal Te of fir=neo) | : +P Gib offer = No) | = 1 x ut x Apo : ag OT 7 oS = 0.0333 | 30Stef: Use martin A poskeriat (mae) by pathasia bar Cynaterr to classify the sagt sbjot—to —the kypotheis with the highest prsbabes tba Since Gib offer yes [Test data) has the hughest- value , Dhe Hest Jeb Offer, = Ys oS Acta is classified as I4- Consider te Sample data given in -lhetable. usith 2 featiaes x and “Y The thaget classes are | Predict he Class UALna- Neasoxt— Cemtacid § claspi fies . Note : Chiven test instance. is (6.5) © pa dice he clos. fA’ or 6B, 7 We COM |Step | 1 Comp rnte the mean |eent rcidl of cach alias Tn Bir example. Whne ane 2 eauccs Aled “A | ond B" . Contnsicd of Class’! = (245+4,14243) =(2 6)/a 3 = (4,2 Contant oa clans = (tote e6t78): (ai, is)fa : “ = (7.6) Neue given a este instance (6,5) tte cam Predict Ate efass. St @: Calculate the Euidean distance bel elieeen ett instance (6,5)° and “each Othe Cent wid. Eve ist [(6.5) 5 i 2»)= (6-44 (5-2) = Viz =36 wist [(65); Gey] = \|(e-7+ (S-6F = Ve ab aig ie Step a. The 4st instance’ has smaltir distanced classy Bo Hemee tte Claas Ais teat instance is predicted ay SB