Linear Algebra Assignment | Fall 2024

⁻ It is required to submit an assignment (15 points).

Please follow the following instructions

 You are required to form teams (5 - 6 students)

 Each team is asked to submit a single report consisting of:

a) Cover page with students’ names and IDs.

b) Detailed answer and Solution procedure for every point of the requirements.

 The discussion will be on Tuesday (31/12/2024) & Wednesday (1/1/2025) & Thursday

(2/1/2025)

 A printed computer-based report (hardcopy) must be delivered during the final discussion.

 You are NOT to seek help from any of your colleagues. You may search online for help but

do NOT copy and claims as your own.

 Zero will be assigned to copied reports.

 Super efforts will be appreciated (Bonus up to 3 marks).

 During the discussion, you may be asked in any part of your report so you must be Aware of

all the steps.

Question [1]: 2D / 3D Geometric Transformation
 Rotation & Scaling & Reflection
a) We need to determine two linear transformations, 𝑭𝟏 (𝑿) = 𝑩𝟏 𝑿 and 𝑭𝟐 (𝑿) = 𝑩𝟐 𝑿, where 𝑩𝟏 is the matrix
that produces the first transformed image from the original, and 𝑩𝟐 is the matrix that achieves the second
transformation from the original image as shown in figure 1.

Figure 1

 Translation & Scaling & Reflection

b) A linear transformation 𝑭(𝑿) = 𝑩𝑿, where B is a transformation matrix, naturally keeps the origin fixed
since 𝑭(𝟎) = 𝟎. However, we can extend this by adding a translation, defining 𝑭(𝑿) = 𝑽 + 𝑩𝑿, where 𝑽 is
a vector. In this case, 𝑭(𝟎) = 𝑽, shifting the origin to 𝑽. For each of the three transformations shown in Figure
2, it is required to determine the distinct vector 𝑽 and the matrix 𝑩 that describe each transformation.

Figure 2
  Shearing
We need to determine the transformation matrix that can describe the following 3D
transformation from the original image into the sheared image that shown in figure 3.

Figure 3

Hint:
Writing code in any programming language to display the original shapes and demonstrate their
transformations would be an added bonus
Question [2-A]: Application Based on Linear System of Equations [Interpolation]
As you move farther out into the solar system and away from the Sun, an object needs to travel at a slower
speed to maintain its orbit. Below are the average orbital radii of the planets in our solar system and their
corresponding average orbital velocities around the Sun.

Planet Distance from Sun (million km) Orbital Velocity (km/s)

Mercury 57.9 47.4

Venus 108.2 35.0

Earth 149.6 29.8

Mars 228.0 24.1

Jupiter 778.5 13.1

Saturn 1432.0 9.7

Uranus 2867.0 6.8

Neptune 4515.0 5.4

a) Find the unique first degree polynomial whose graph passes through points defined by Mercury and Jupiter.

b) Plot the data points together with the graph of the polynomial to observe the fit.

c) Adjust your polynomial and graph by adding Saturn, and then Earth. What do you notice as you add more

points? What if you had started with different planets?

d) Expand your work in part (a) to a seventh degree polynomial that passes through all eight planets.

e) The first object in the Kuiper Belt, Ceres, was discovered by Giuseppe Piazzi in 1801. Ceres has an average

distance from the sun of 413.5 million km. Based on the points on the graph, estimate the orbital velocity of

Ceres. What does the polynomial suggest the value would be?
Question [2-B]: Engineering Application Based on Linear System of Equations
Consider a thin square plate whose faces are insulated from heat. Suppose that the temperature along the four
boundaries of the plate is known, and further suppose that those temperatures are held constant. The following
diagram represents this situation.

The temperature inside the device will reach certain equilibrium after
some time has passed. consider a few points on the plate
and approximate the temperature of these points, this
approximation is based on a very important physical property called
the Mean-Value Property. By the second version of the Mean-Value
Property, we can express the temperature 𝑥1 at point 1 by the
following equation:

𝟏𝟎 + 𝟑𝟎 + 𝒙𝟐 + 𝒙𝟑
𝒙𝟏 =
𝟒
 Follow this representation of the temperature of each point to convert the given diagram into system of
linear equations. (write the system in the matrix form 𝑨𝒙 = 𝒃 where, the vector 𝒙 is the vector of
equilibrium temperatures).
 Prove that the matrix A in this system is invertible and compute its inverse then find the vector of equilibrium
temperatures which represents the temperature in degrees Celsius at the four points inside the device.
 Suppose that the boundary temperatures change from 30, 10, 0 and 20 to 15, 10, 10 and 20 respectively.
What is the effect of this change on the Temperature Distribution linear system?
 A finer grid should give a better approximation of the equilibrium temperatures. Consider the grid in the
following Figure and find the vector of equilibrium temperatures. Use any programming language to find
the vector of equilibrium temperatures.

Hint
Using any finite element software such as COMSOL Multi-physics,
ANSYS, Matlab, Autodesk CFD…. Etc. to compare the results will be an
added bonus
Question [3]: Graph Theory & Network Analysis

Figure 4
An adjacency matrix is a square matrix used to represent the used to represent a graph. The rows and columns of
the matrix correspond to the nodes of the graph, and the matrix entries indicate the presence edges between
nodes such that:

 1 if there is an edge between two nodes.

 0 otherwise.
It is required to do the following tasks:
a) Build the 7 × 7 adjacency matrix that represent the given network shown in figure 4.
b) Search how an adjacency matrix can be applied to find the shortest path length between any two nodes in
the network and how it can be used also to determine the number of possible ways to reach one node from
another one with a specific path length.

c) Hence, find the shortest path length to move from Node 3 to Node 7 then return back to Node 3. Find also
the number of possible ways to do that.

A clique in graph theory is a subset of vertices within a graph such that every
two distinct vertices in this subset are directly connected by an edge. For
example, as shown in figure 5, nodes 2,3 and 4 forms a clique.
d) Explain how adjacency matrix can be used to find the cliques in the
network.
e) Hence, use this to find and highlight the cliques in the given network
shown in figure 4.
Figure 5
Question [4]: Choose one of the following topics to search about it.
1- Encryption Using Enigma Machine

The Enigma machine, which was developed by a German engineer, played a crucial role in encrypting German
messages during World War II. This cryptographic device, relying on both electrical and mechanical operations.
The enigma machine encrypts each letter by three processes giving a total of 7 encryptions operations for each
letter and each process was a carried by a different component.

 Your task is to search about Enigma Machine and know its components and the rule for each part.
 Hence, write a code using any programming language that simulates the operation of the Enigma Machine.

2- Facial Recognition Using Eigen Faces & PCA Algorithm

Face recognition is a biometric technique used to identify or verify a person by analyzing their facial features.
Principal Component Analysis (PCA), is a statistical method that reduces the dimensionality of the data while
preserving the most critical features. PCA identifies patterns in the data and transforms face images into a lower-
dimensional "Eigen face space," where each face is represented by its unique combination of principal
components.
Steps in Face Recognition Using PCA:

1. Data Preparation: Collect and preprocess face

images.
2. Mean Face Calculation: Subtract the average face
from all images to normalize the data.
3. Feature Extraction: Use PCA to identify the principal
components (Eigen faces).
4. Projection: Project face images into the Eigen face
space.
5. Classification: Compare the projected images to
classify or recognize faces.

 Your task is to write a code using any programming language that applies face recognition using PCA and
Eigen Faces algorithm.