ASSIGNMENT-2
Q.1. Construct a Linear Regression Model for the following data set. Predict the price for a house with area
2200 sq ft and 3 bedrooms.
Size (sq ft) Number of Bedrooms Price (in $)
1000 2 300,000
1500 3 450,000
2000 3 500,000
2500 4 600,000
3000 4 700,000
Q.2. Construct a Linear Regression Model for the following data set. Predict the price for a car with age 3 years
and mileage 50,000 miles.
Age (years) Mileage (miles) Price (in $)
1 10,000 25,000
3 30,000 19,000
5 50,000 15,000
7 70,000 12,000
10 100,000 8,000
Q.3. Actual and predicted values are given for 10 students in the following data set.
STUDENTS ACTUAL PREDICTED DATA
DATA
1 PASS PASS
2 PASS FAIL
3 FAIL PASS
4 FAIL FAIL
5 PASS PASS
6 FAIL FAIL
7 PASS PASS
8 PASS FAIL
9 FAIL PASS
10 FAIL FAIL
(a) Find the recall and precision for the above dataset.
(b) Find the F-1 score and confusion matrix for the above dataset.
Q.4. Use Logistic Regression to predict the result (pass or fail) of a student based on number of study hour. Use
the model to predict whether a student will pass if the number of study hours is
(a) 4 (b) 6
Hours Studied Passed (1 = Yes, 0 = No)
1 0
2 0
3 0
4 1
5 1
6 1
7 1
𝟎, 𝒙 < 𝟎. 𝟓
Use the Threshold Function as 𝑻𝒉𝒓𝒆𝒔𝒉(𝒙) = { and 𝒘 = (𝒘𝟎 , 𝒘𝟏 ) = (−𝟔, 𝟏. 𝟓).
𝟏, 𝒙 ≥ 𝟎. 𝟓
Q.5. Construct a Decision Tree for predicting whether a student will pass an exam based on the features: Study
Hours and Number of Homework Completed.
Study Hours Homework Completed Passed Exam (Target)
A1 B2 0
A2 B1 0
A2 B3 1
A3 B1 0
A3 B2 1
A4 B4 1
Q. 6. Suppose we have the following dataset. In this dataset, there are four attributes. And on the basis of these
attributes, make a Decision Tree.
Age Competition Type Profit
Old Yes software Down
Old No software Down
Old No hardware Down
Mid yes software Down
Mid yes hardware Down
mid No hardware Up
mid No software Up
new yes software Up
new No hardware Up
new No software Up
Q.7. Derive the Mean Squared Error function in matrix form for the Linear Regression Model, i.e. Derive
1
𝐿(𝑤) = (𝑋𝑤 − 𝑦𝑡𝑟𝑢𝑒 )𝑇 (𝑋𝑤 − 𝑦𝑡𝑟𝑢𝑒 ).
𝑚
Q.8. Derive the minimum value for the MSE 𝐿(𝑤) using gradient. i.e. Derive 𝑤 = (𝑋 𝑇 𝑋)−1 𝑋 𝑇 𝑦𝑡𝑟𝑢𝑒 .
Q. 9. Use K-Means Algorithm (up to the second iteration) to cluster the following eight points into three
clusters: A(2, 10), B(2, 5), C(8, 4), D(5, 8), E(7, 5), F(6, 4), G(1, 2), H(4, 9) with A(2, 10), D(5, 8) and G(1, 2)
as the initial cluster centroids. Consider the following two distance functions in two different cases.
(a) The distance function between two points 𝑎 = (𝑥1 , 𝑦1 ) and 𝑏 = (𝑥2 , 𝑦2 ) is defined as 𝑑(𝑎, 𝑏) = |𝑥1 − 𝑥2 | +
|𝑦1 − 𝑦2 |.
(b) The distance function between two points 𝑎 = (𝑥1 , 𝑦1 ) and 𝑏 = (𝑥2 , 𝑦2 ) is defined as 𝑑(𝑎, 𝑏) =
√(𝑥1 − 𝑥2 )2 + (𝑦1 − 𝑦2 )2 .
Q.10. The dataset of pass or fail in an exam of 6 students is given in the table.
Study Hour 29 16 33 28 29 30
(X)
Pass (1), 0 0 1 0 1 1
Fail (0)
Use logistic regression as classifier, calculate the probability of pass for the student who studied 30 hours, where
𝒘 = (𝒘𝟎 , 𝒘𝟏 ) = (−𝟔𝟒, 𝟐).
