Scribd
Upload
56 views25 pages

DWM Module 3

The document discusses various classification methods, including Decision Trees and Naive Bayes, and their applications in predicting outcomes based on input data. It emphasizes the importance of evaluating classifier accuracy and provides examples of how to calculate probabilities and entropy in decision-making processes. Additionally, it touches on the holdout method for training and testing classifiers using datasets.

Uploaded by

vv376012
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
56 views25 pages

DWM Module 3

The document discusses various classification methods, including Decision Trees and Naive Bayes, and their applications in predicting outcomes based on input data. It emphasizes the importance of evaluating classifier accuracy and provides examples of how to calculate probabilities and entropy in decision-making processes. Additionally, it touches on the holdout method for training and testing classifiers using datasets.

Uploaded by

vv376012
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
Poa $$ ah Explodn Holdout—and_ fonden_—eubsenp lif method ty Jevaludte the acceya of classifier LE: 5. . Guive the draining. be bls, 3 classi the tuple, t==go then prode A if ope xen go Pingrede & | 4 od=nc=8* then grade © tif 662=% <= 70 then prade D | th xe so the grade € 1 A _|Classificatinn Metheds: = 22 1|Decision Tree Taductlon \- 4} Boyesion classificablyn Naive Bayes classifier — “I Decision Tree Inductisn is the" process. of- audamadially building a decision Five trom: "-a.él6ss ~ labeled | | ty dadasebs f eb ah ied “The "ids wise” ned az ge tae a piste [where = Back =xecovd ole outs alwys Taye T#erati vo” Tichadaisey $ CHS ~ Successor af: ID 3.\ 3. CART - ee iftcahlen and Kesreaslon Tree. 4; s CHAT = Ghi_- Automatic: Ti; bn Dé ra guave Que. || Explas \slon Tie. A Dewision Tree. js* a It ts represented “ds. a: dee like — where: ||_- Tateraal node apres teat an_attributes 22 -|+ Byanches. SST a eoutenmies of hase tests. tenet Si the “Deck Lon Tree Based. Cla Ticollan spate ——_|_cuperyised learning. _techingue because __'+ uses _o leboled deg Geach +ts__learn__the_decisisn syules classmate aa Ste Paes a “Decisian: Tree set ee su selett He heat aie fo: ae iy fusing eth = at “mie ciwhnd. £ Sbleolel, strong Table. seta speed to diad. hich abfathake alll be the jest pedercin dur deelslen tree the “gain ts coleutated fer all tour ashy ca formula th -opalns 2 oct anbsssig alt + i we Se ee eascoate page = —~ | sample cedesetg Great - i \ [Day| outlook | Temperatore| Humidgty | Wind __| Play ball | sunny Wet | dish | seat No hat $1 Lsuniny: cio. Hoth saith ~ Lstrong! =: bs Was 7 overcast | __Hot _ | High leak} tes _ 4 | Rain = Mild Hip Weak Yes S| hala cool Norma! | weak Yes 1:6 | Pein. coo! ‘Norma Strong. No ‘| 7 lQvevmat | Cool Normal ston Yes la Mild "| High - No ls Cools Noymal- “en | Yes Jo ot Mile © | Namal weak 9 >|. Yes. : att Mild Nlwrmal_|_strong Yes 12 ft Mild > | High. -basdrong. 4 Yes 13 |Qyercast |__ Hat Norma! | weak Yes 1 Rain Mild ‘ Hh etter No Solp "| Caleulate Frbmalie. Gale (568 ob outlook — is Noob Ayes= g Jxic oe 8 Fee ' Noof nto= S Steps? Heph lg Eopyb—eae_chhs a Tie Spam Slog oo WA AS yee og gg 14 sip eaop of- al) cathcibutes. : » Enhopy F sung, @2, at = lag 2. - 3 lag classmate, — sEndeopy of fain $5.23 223 yg Si = 2 ge ath hat ov slat iseetrosg pet 2 pigk7 aN oust | ~Galeulate “the TG..of Temjenafure i= © - lo. = “2. = Giopy mote a he hc ep. Mild St, 25 oe tet er classmate, a Enttopy of - a abla set = 0.94 + Zalogs at all Fie odiudes ; a ee ee = 0:38 i Exyy ot Numi £03. = =$ t,t 5 yp § A fen fone | = 0+ 032 1 aT. yk at \ wllsaral +t peal te Teno tantly | — m 7 enn) —| aa3 toes =0t Ake nant = 0-13 + Coleulate the TG. ob lind Enkopy of Whole slute set = 0.94— Eat of-_all the ottiibufes = ap beak ot bi hE 0-8) eo ee classmate q Eninpy ob dary § 3/83 = = leg op re fone nO TabspmahionGrala_= Evdapy (whole elala)—6 Entic) 8&_ Ent yr tlele siplat ‘ie [tot =O. 94- = 8 eattyogy : og oa - VE arakad =O: 047 inal Grains * - Crain (S tut eak PPh oa. Cn = 082 ‘Gain (Ss Humidita) = 0.15 ~ Gain {s, wind) Ve 0.047 pion ee a eo As outlook. has’ oats rahued * [Link] vain", the peek a has three Gere a Sukny oveycast » Rosln * ¥ rok Tempered eb |_ al | talind ey bal) pl = Her. js High Sf Add eke ts No Hot E High Strong. No Mile | bid | hea No Coo) Norma) | weak | Yes Mild >| Aldcmascl sheng. | Yes | Caloulate sila cree nied py ot- Und $2, BE eB 2 toate — a = O97 LEatopy ofall the attribules ! Enkespy od Hof {6,2% = Enkeepy ok WIM PAE Se pe a A\Loh, osinpy tes Gn oh =o Ld, i iQ ee Enhopy “Ceammyle 2 B eolo)~ ene) | —Lemtes | ZI I" o Co Fa au ‘a t > n ie 7 la KI Loy classmate 2 ras 3 — C) | = = see as = 0. ———s Ee ee ee ee Entropy ode: Neca | fo,0 =0 E Entosmapien - EL Feadenpy Crain = Entropy (sunny). 3: Ent(Hl= 2 cob (ayl = 69] —-60-0 a BOSE ee eh Va ke eda | | Coloulote the “TG of-t Wind’? 2 ( Entropy = of wang LBV BF = 0-37. t I 1 Entropy of ae futiety iged = Z =O. als pep - Enteopy o4- adeany XL R= ho : Tadecxonahon_geuin > Entropy | (sunny) 2 oi - = ent-(s \ | i ily Tejeda) = 8 at Ack Ne (Banay “Wind = 10:02: set | Humnichity jah ge therefore, $b e- aT elt = nfl} jes eBeaeine Aly of every yen, al om nt to coleulabe aryethlip— assyoed “yes JouHtook. | syn aes reas Yes ih None] —— wet Nyes T . Day| outlook | Temperature | Humidity |- Mind | Playball ! ral Rou Mild he jeleak Ye. ! IS | pal G Noyme)] leak |. -> Weg 16 | tain Cool | bthehtrml sfrony | No 1s_| Rody Mild v2) Nema). |. weak =| ~ Yes | 1 | aln Ne we eMild © th “Sheng | O:-019 OF 9}- Bx 08 516 =e, ry Entropy oe foun} 3,23 2 O97 classmate. Si \\—_——— [Eahopy of all “atteibupe __Enteopy a |_| 1 Enitepy. vob Normal 42.1% =-2. 2g en A 3°03 se = 0.918 | l= TE esp) — f | = O47 - st-4 Log |: = 0019 H | Calewlade the Gro —bedind. Endopy of — Weak 4 108 so | Entropy SZ sing 2 25 =0 fem 2 0-97 Hence. Crain (Span Temperate) = 0019 4 fyes) P(>Fo) No)= 2 wwe} frachice Mere Westiors. Emp. ff elascrate | Je 0) | Sheaiunl yes) ; plneglimlyes)= & C ello |-yes) =e ete a #lyeal ves) , = IC aL “nila Li} Ee c a = C A exclert 0) =. Pevideal Tye ook : pl ye \ “ptney 2S | Te — ; = fhe: wall zi ‘ pe student= thes", cvedit rating =" calets | unseen ample Ae ages! "4 30'[Link] ~'medllun” ‘e(x} yes}. Plyes) = plage= "250 oT yes x p(Intome>" redo) ar Hetil "es" Ye) "pl edhdy= *Falv'| Yes) pl tes} — a Z Ne ee: | SG Gb — Tig ome = viet 7 classmate |p ela see) 5 oes, 36" wade ed “meetium’| No) — = : tidents "Yes" He) x Pletal " ; TT : ~ x Ali) : “peachy 4s | 1 Jace: 0.82.8 > 0.6068. , There for jhe naive Basin. classifiey eae yes" fy a. x i Holdout |! Tn__holdau, a ya ce tay — ‘yo dade eet ond test clote cet Cusualles 732-48 testy 213 for, sealing i : ToFal_ Namber of: examples : | Training Seb | *TearSer fa — || To tratn__the- _classifier ays tad dotq__sef-_useol | and__ance___the classifier is” constructed then use dest _dhta__sot to estimate the _evnr. vote of fhe cla. ifier. ' : daset Ja Mavgei’ the" fad + Ibe.

You might also like