You are on page 1of 4

CSE 317 NAME :

Final Exam ID :
Time 1 hour 30 mins SECTION :
Total points: 45

Answer any 3 of the following questions.

1. [CO1] [15 points]


a. [5 points] Show that the function given below is not linear.

b. [5 points] Find the 1st order Taylor approximation of the above function around .

c. [5 points] Find the approximate value of the function at the point using the Taylor
approximation.
2. [CO2] [15 points] Consider the vectors .

a. [2.5 points] Show the vectors form a basis of the space

b. [5 points] Find a unit vector orthogonal to both vectors. (You cannot use the cross

product.)

c. [2.5 points] Show the vector is not in .

d. [5 points] Find the vector in that is closest to in the least square sense.
3. [CO2] [15 points] Consider the least square regression problem with two features/attributes. In the
table below we present 4 data points. and are the features and is the output variable.

x1 x2 y
1 3 14
3 0 9
0 1 8
1 5 25

a. [10 points] Suppose we want to predict using only the feature . Find the coefficients of the
regression function by solving the normal equations in matrix form using Gaussian elimination.

b. [5 points] Write down the corresponding system of linear equations in matrix form for the regression
problem when both the features are used. Write down the corresponding normal equations.
4. [CO2] [15 points]
a. [10 points] Find the decomposition of the following matrix using Gram-Schmidt
orthogonalization.

b. [5 points] Describe the process to obtain the least square solution of a system of linear equations
using decomposition.

You might also like