Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Pir Mehr Ali Shah

Arid Agriculture University, Rawalpindi


Office of the controller of Examinations
Final Exam (Theory) / Spring 2021 (Paper Duration 12 hours)
To be filled by Teacher

Course No.: ……CS-632 ……Course Title:…………Artificial Intelligence …………………


Total Marks:……20 …………Date of Exam:…………9-7-2021.....................
Degree: ……BSCS…………….Semester:…6 ……… Section:………A/C/D…………………………
Marks
Q. No. 1 2 3 4 5 6 7 8 9 10 Obtained/
TotalMarks
Marks
Obtained
Total Marks in Words:
Name of the teacher: Saba Un Nisa
Who taught the course: Signature of teacher / Examiner:

To be filled by Student

Registration No.: …………18-Arid-5157………….……… Name:……Ibtasam Mateen Javed…………………..

Answer the following questions.

Q1 Using Decision Tree algorithm on the following dataset to draw ROOT node only.
(Where fast is output attribute)

Answer:
S= Fast= [yes, no]
S1= [5+ , 10-]
Entropy(s)=-5/15 log25/15-10/15log210/15
=0.333+0.3899 = 0.7232
Values(Engine)= [Small, Medium, large ]
SSmall [ 1+,5- ]
SMedium [2+,3- ]
Slarge [2+,2-]
Gain (S1 Engine)= Entropy(S)-∑|Sv|/|S| Entropy(Sv){ Small, Medium,large,}
Gain (S1 Engine)= Entropy(S)-6/15 (SSmall )-5/15 Entropy(SMedium)-4/15 Entropy(Slarge )
Entropy (Small)=-1/6 log21/6 – 5/6 log2 5/6 = 0.65
Entropy (Medium)=-2/5 log2 2/5 – 3/5 log2 3/5 = 0.97
Entropy (Large) = -2/4 log22/4 -2/4 log2 2/4=1
Gain (S, Engine) =0.7232-6/15*0.65-5/15*0.97-4/15*1 = -0.1268
Values( SC/Turbo)=[Yes,NO]
Syes  [2+,2-]
SNO[3+,8-]
Entropy(Syes)=-2/4 log22/4 -2/4log22/4=1
Entropy(SNO)=-3/11log23/11-8/11log28/11=0.8453
Gain(S, SC/Turbo)=Entropy(S)-4/15 En(Syes) -11/15 En (Sno) = 0.7232-4/15*1-11/15*0.8453=
-0.163
Values(Weight)=[average , light , heavy ]
Saverage [3+,3-]
Slight [1+,3-]
Sheavy [1+,4-]

Entropy(Saverage)=-3/6log23/6-3/6log23/6 =1
Entropy(Slight)=-1/4log21/4-3/4log23/4=0.811
Entropy(Sheavy)= -1/5log21/5-4/5log24/5=0.721
Gain(S,Weight)=Entropy(S)-6/15En(Saverage)-4/15 Entropy(Slight)-5/15 Entropy(Sheavy)
=0.7232-6/15*-4/15*0.811-5/15*0.721=-0.1334
Values(Fuel Eco)=[Good,avarge,bad]
Sgood [0+,2-]
Savarge [0+,3-]
Sbad [5+,5-]
Entropy(Sgood)=0-2/2log22/2=0
Entropy(Savarge) =0
Entropy(Sgood)= -5/10log25/10 -5/10log25/10=1
Gain(S, Fuel Eco)=Entropy(S) - 2/15En(Sgood)-3/15 Entropy(Saverage)-10/15 Entropy(Sbad)
=0.7232-2/15*0-3/15*0-10/15*1= 0.056
So, according to the information Gain measure, the Fuel Eco attribute provide the best
prediction the output attribute fast,
Root Node  Fuel ECO

Q2. Apply the Alpha Beta Pruning Algorithm on the Following given tree?

Answer:
Q4. Consider the below given iris data set. This dataset contains 3 classes of 15 instances each
and each class refers to a type of iris plant. The dataset has two features: sepal length, sepal
width. The third column is for species, which holds the value for these types of plants. A new
plant is identified. You have to classify the Species class of new identified plant with the help
of KNN algorithm and K= 3.

Answer:

First Step to find the Euclidean distance between the Observed and actual length and width
X =Observed Sepal length =5.2
Y= Observed Sepal width = 3.1
A= Actual Sepal length =5.3
B= Actual Sepal width =3.7
_________
Distance =√(𝒙 − 𝒂)𝟐 + (𝒚 − 𝒃)𝟐
_____________
=√(𝟓. 𝟐 − 𝟓. 𝟑)𝟐 + (𝟑. 𝟏 − 𝟑. 𝟕)𝟐 = 0.608
Sepal length Sepal width species Distance
5.3 3.7 setosa 0.608
Now for second row
_________
Distance =√(𝒙 − 𝒂)𝟐 + (𝒚 − 𝒃)𝟐
______________
=√(𝟓. 𝟐 − 𝟓. 𝟏)𝟐 + (𝟑. 𝟏 − 𝟑. 𝟖)𝟐 = 0.707
Dis for third row Now for distance for all rows and the table is as follow
Sepal length Sepal Width Species Distance Rank
5.3 3.7 Setosa 0.608 3
5.1 3.8 Setosa 0.707 6
7.2 3.0 Virginiea 2.002 13
5.4 3.4 Setosa 0.36 2
5.1 3.3 Setosa 0.22 1
5.4 3.9 Setosa 0.82 8
7.4 2.8 Virginiea 2.22 15
6.1 2.8 Virginiea 0.94 10
7.3 2.9 Virginiea 2.1 14
6.0 2.7 Virginiea 0.89 9
5.8 2.8 Virginiea 0.69 5
6.3 2.3 Verscicolor 1.36 12
5.1 2.5 Verscicolor 0.60 4
6.3 2.5 Verscicolor 1.25 11
5.5 2.4 Verscicolor 0.75 7

Find the Nearest neighbor

For K =1

Sepal length sepal Width Species Distance Rank

3.1 3.3 Setosa 0.22 1

Feature Species is Setosa so for K=1 it is setosa

Sepal length sepal Width Species Distance Rank

5.1 3.3 Setosa 0.22 1

5.4 3.4 Setosa 0.36 2

Feature Species is Setosa so for K=2 it is setosa

For k=3

Sepal length sepal Width Species Distance Rank

5.1 3.3 Setosa 0.22 1


5.4 3.4 Setosa 0.36 2

5.1 3.9 Setosa 0.608 3

Here

NO of Setosa are =3

No of verscicolor are =0

No of virginica are=0

So majority is Setosa

Final Answer

Sepal length sepal Width Species

5.2 3.1 Setosa

Q3. By using single layer perceptron algorithm classify the following patterns. Where X1
and X2 are inputs Y is the output. Learning rate is 0.5 , output is 1 if weighted sum >=0.5
otherwise 0 and weights w1 =0.2 and w2 = 0.3
X1 X2 Y
0 0 0
0 1 1
1 0 1
1 1 1

Answer:

For case 1 & case 2 no error is generated

For case 3 X1 = 1 , X2 = 0

W1 X1 + W2 X2 = 0.2x1 + 0.7x0

W1 X1 + W2 X2 = 0.2

F(W1 X1 + W2 X2 ) = f(0.2) = 0
but actual output is 1.

dԐ dԐ
For W1 = W1 .α. W 2 = W 2 ..
𝑑W1 𝑑W2

dԐ dԐ
= X1 + (Error) = X2 + (Error)
𝑑W1 𝑑W2

W1 = W1 – α. X1 (0 – 1) W2 = W2 – α . 0 (0 – 1)
W1 = W1 – (0.5). 1 (– 1) W 2 = W2

W1 = W1 – (0.5)(– 1)

W1 = W1 + 0.5

After Updating Weight.

W1 = 0.2 + 0.5 = 0.7

W2 = 0.3

Now when we ran all core the error will be zero.

Case 1:-

0, 0 = 0.7 x (0) + (0.3) x 0 => f(0) => 0

Case 2:-

0, 1 = 0.7 x (0) + (0.3) x 1 => f (0.3) => 1

Case 3:-

1, 0 = 0.7 x (1) + (0.3) x 0 => f (0.7) => 1

Case 4:-

1, 1 = 0.7 x (1) + (0.3) x 1 => f (1) => 1

Therefore final weight are

W1 = 0.7

W2 = 0.3

You might also like