Question

4. Reconsider the tennis playing training examples, if B Bayesian Belief Network depicting the conditional independence relationship between the attributes and target classification are shown as follows: Day Outlook Temperature Humidity Wind Play Tennis gh Weak No Sunny Ho gh Sunny Ho Strong No gh Weak Yes Overcast Ho Mild gh Weak Yes Rain Normal Weak Yes Rain Coo Normal Rain Coo Strong No Normal Strong Yes Overcast Cool Mild gh Weak No Sunny Normal Weak Yes 9 Sunny Coo Normal Weak Yes 10 Mild Rain Normal Strong Yes 11 Sunny Mild 12 Overcast Mild gh Strong Yes Normal Weak Yes 13 Overcast Ho 14 Mild gh Rain Strong No Outlook Humidity Wind Temperature Play Tennis (a) Initialize the following conditional probability tables (CPT) with respect to the training data? Outlook Sunny Overcast Rain Outlook Temperature Shot Temperature mild Temperature cool Sunny Overcast LOutlook Humidity hot Humidity normal Sunny Overcast

Answer 1

Total Tenere are 14 instances و outlook | Rain Sunny I overcast outlook Temp= cool | Tempahot Tem=mild Sunny مران از - ا overcast اس لالو کار Rain و می

outlook | Humidity=high H Humidity = normal alo llo sunny overcast Rain hole slupla da solo Temperature, Humidity play=yes I play=no n Hot High alm Hot Normal mild, High Mild, Normal coll, High coll, normal op dielony of all oport normal

PLYES) - 5 9 Tú P (Nol= pf windweak tyle having instances inces Humidity = Hight num of Humidity = normal =) instances having the probabibity for Humudity High num of so n 7 wormals in P ( Highlyes) = plnes/High) xP (High) PLYES

P ( High / No) = P(Nol High) P (High) P (NO) Pirosmal lygs) - PC Esimorvial] * p loov mal) Plyes) p(normal) P(Namal (No) = P(Nol Normal) x el P(NO) 나 HELE 아

num of instances having Tema Hot =4 Plres/ Hot ) a num of instances having Tema Milda G P(yes/ Mild) = 4 mium of instances having Tem= cool = 4 Plyes/ cool) = 3 P (Hotlyes) =P(YES | hot) app hot) P(Horino) = ax w Plyes dit Han dat dit all fort (Mild Inol P( mildlyes) upl res| mild & P (Mild) Plyes

미 1001yes) =p[yes \cool) x PC cool) P4 회에티지 M/

outlook - 3 ok = sunny=5 P(Ves / sunny num of instances having a - overcast-4 P(yes lovely - in of instances having outlook outlook Rain-5 Plyes/ Red having am of instances P(sunny lyes) = P(yes/sanny) *P (sunny) P(sunny/No) = alls P(Yes 7/9/519 - 1 x Ty = 14 als p(overcast No) =0 X p(overcast lyes) = P(yes lovercast it holt Ho

* P(Rain) ܠܐ ܘܐܠܘܠܣܪ et p( Ranival yes) = P( res a 9 x foutlook-sunny temperature = shot humidity srovnal p( xl yes) = f(kes) v panche) - (notlpes+ Pral|yed

! PL xl yes) = m(Yes) < p sunettes) xp hot (yes) 21x9 P(ANO) = P(NO) to some hos) * pfhot (No) te p (normal wo) ook 9.29 so the play tennis is NO