DEMO VERSION OF THE SECOND ROUND TASKS

Computer and Data Science: Second-round Sample Tasks for the

Open Doors Postgraduate Track

You will be asked to complete 35 tasks, including:

- 21 entry-level tasks, each correct answer worth 1 point;
- 10 intermediate-level tasks, each correctly answered task worth 3 points;
- 1 intermediate-level task worth 4 points;
- 3 advanced tasks (constructed response), each correctly completed task valued at 15 points.
Evaluation criteria and standard answers are provided for the advanced tasks requiring constructed
responses.

Applied Mathematics

Task 1
Entry level (1 point)

In accordance with the POSIX Extended writing standard, the following regular expression is
/^[abc]+[a-zA-Z]?8{3}/. Select the string that corresponds to it.
a) aaFDF888
b) aaF888
c) xzyaaR888
d) aaR8
Answer: b.

Task 2
Entry level (1 point)

If the graph is given by an adjacency matrix:

{0, 1, 1,0,1 1,0,0,1,0 1,0,0,1,1 0,1,1,0,1 1,0,1,1,0 },
What is the weight of the minimum spanning tree?
a) 2
b) 3
c) 4
d) 1
Answer: с.

Task 3
Entry level (1 point)

Which Boolean function does this scheme of functional elements correspond to?

1
DEMO VERSION OF THE SECOND ROUND TASKS

a) xw ⋁ xz
b) xy ⋁ wz
c) xw ⋁ xz
d) xw ⋁ xz
Answer: d.

Task 4
Intermediate level (3 points)

Which of the following functions is in disjunctive normal form (DNF)?

a) f (x, y, z) = x⋀x ⨁ x⋀y⋀z
b) f (x, y, z) = (x⋁x) ⋀(x⋁y⋁z) ⋀ (y⋁x)
c) f (x, y, z) = x⋀y⋁ x⋀y⋀z
d) f (x, y, z) = x ⋁ y ⋁ z
Answer: c, d.

Task 5
Intermediate level (3 points)

For which sets A and B, is |A ∪ B| = 5?

a) A = {1,2,3} B = {1,6,7,2}
b) A = {1,2,3} B = {1,6}
c) A = {1,2,3} B = {6,7}
d) A = {1,2,3} B = {1,2}
Answer: a, c.

Task 6
Advanced level (15 points)

Transform the function F(x1 , x2 , x3 , x4 ) = 1001 0010 1010 1100, using the formula of
disjunctive decomposition over a set of variables х1, x2. Provide the resulting functions for the
variables x3 , x4 , using a set of operations from the basis: NOT, OR, AND.

Please note that the evaluation will consider your problem-solving process; providing only the
final answer is not sufficient.

2
DEMO VERSION OF THE SECOND ROUND TASKS

Solution:
Using the disjunctive decomposition formula for two variables, we obtain:
a a
F (x1 , x2 , x3 , x4 ) = ⋁(a1 ,a2 ) x11 ∙ x22 ∙ f1..4 (a1 , a2 , x3 , x4 )
a a
= ⋁(0,0) (0,1) (1,0) (1,1) x11 ∙ x22 ∙ f(a1 , a2 , x3 , x4 )
= x10 ∙ x20 ∙ f1 (0,0, x3 , x4 ) ∨ x10 ∙ x12 ∙ f2 (0,1, x3 , x4 ) ∨ x11 ∙ x20 ∙ f3 (1,0, x3 , x4 ) ∨ x11
∙ x12 ∙ f4 (1,1, x3 , x4 )
= x1 ∙ x2 ∙ f1 (0,0, x3 , x4 ) ∨ x1 ∙ x2 ∙ f2 (0,1, x3 , x4 ) ∨ x1 ∙ x2 ∙ f3 (1,0, x3 , x4 )
∨ x1 ∙ x2 ∙ f4 (1,1, x3 , x4 )
(A correct entry of the disjunctive decomposition formula for the specified variables is
worth 5 points.)
Let us determine the values of the functions f1, f2, f3, f4 on each set, using the values on
similar sets of the original function F (you can use the truth table or the Karnaugh map for visual
comparison, which are presented below). For example, in the first column, we alternately define
f1 (0000) = 1,
f1 (0001) = 0,
f1 (0010) = 0,
f1 (0011) = 1; we also determine the functions f2−4 .
The truth table of the function F
x1 x2 x3 x4 F
0 0 0 0 1
0 0 0 1 0
0 0 1 0 0
0 0 1 1 1
0 1 0 0 0
0 1 0 1 0
0 1 1 0 1
0 1 1 1 0
1 0 0 0 1
1 0 0 1 0
1 0 1 0 1
1 0 1 1 0
1 1 0 0 1
1 1 0 1 1
1 1 1 0 0
1 1 1 1 0

A Karnaugh map for F

x1 x2 00 01 11 10
x3 x4
00 1 0 1 0
01 0 0 0 1
11 1 1 0 0
10 1 0 0 1

3
DEMO VERSION OF THE SECOND ROUND TASKS

x3 x4 f1 (0,0, x3 , x4 ) f2 (0,1, x3 , x4 ) f3 (1,0, x3 , x4 ) ∨ f4 (1,1, x3 , x4 )

0 0 1 0 1 1
0 1 0 0 0 1
1 0 0 1 1 0
1 1 1 0 0 0

(A correct comparison with the values of the original function is worth 5 points.)
All that remains is to compare each function with the set of Boolean algebra operations,
obtaining:
F(x1 , . . x4 ) = x1 ∙ x2 ∙ (x3 ∨ x4 ) ∙ (x3 ∨ x4 ) ∨ x1 ∙ x2 ∙ (x3 ∨ x4 ) ∨ x1 ∙ x2 ∙ x4 ∨ x1 ∙ x2 ∙ x3
(A correctly written disjunctive decomposition formula for x1 и x2 with pre-defined
functions for x3 and x4 is worth 5 points.)

Computer Science: Information Systems

Task 7
Entry level (1 point)

Which of the following is a telecommunications device?

a) Network adapter
b) Processor
c) Printer
d) Video card
Answer: a.

Task 8
Entry level (1 point)

List the disadvantages of Gantt chart data analysis.

a) inflexibility, dependence, overflow
b) inflexibility, independence, overflow
c) flexibility, dependence, overflow
d) flexibility, independence, overflow
Answer: a.

Task 9
Entry level (1 point)

Which knowledge representation model aligns with the understanding of long-term memory
organization in humans?
a) Production representation
b) Semantic networks
c) Frames
d) Integral representation

4
DEMO VERSION OF THE SECOND ROUND TASKS

Answer: b.

Task 10
Intermediate level (3 points)

Which are the main stages of working with data in information systems?
a) Extraction and transformation stage
b) Machine learning stage
c) Loading phase
d) Data model development stage
Answer: a, c.

Task 11
Intermediate level (3 points)

Which of the following is the purpose of Data Storage (DS)?

a) Ensuring the collection, storage and quick access to the key information
b) Data visualization
c) Operational analytical processing and data mining
d) Viewing video files
Answer: a, c.

Computer Science: Artificial Intelligence

Task 12
Entry level (1 point)

What happens when we introduce more variables to a linear regression model?

a) Both might increase or decrease depending on the variables introduced.
b) The R-squared value may increase or decrease, but the adjusted R-squared always
increases.
c) Both R-squared and adjusted R-squared always increase.
d) The R-squared value may increase or remain constant; the adjusted R-squared may increase
or decrease.
Answer: d.

Task 13
Entry level (1 point)

Which of the following represents the upper and lower bounds of MCC in case of binary
classification?
a) (0,1]
b) [1,0)

5
DEMO VERSION OF THE SECOND ROUND TASKS

c) [-1,1]
d) (-1,1)
Answer: c.

Task 14
Entry level (1 point)

What happens during neural network training?

a) The number of layers and the number of neurons in each layer are adjusted.
b) The parameters of layers and weights are adjusted.
c) The length of the gradient descent step is adjusted.
d) The input object is configured.
Answer: b.

Task 15
Intermediate level (3 points)

Which TWO of the following clustering algorithms suffer from the problem of convergence at
local optima?
a) K-Means++ algorithm
b) EM algorithm
c) Brute force algorithm
d) K-Means algorithm
Answer: b, d.

Task 16
Advanced level (15 points)

Consider a two-layer perceptron with 2 neurons in the hidden layer and 2 neurons in the output
layer. The input vector has a length of 2.
describesis

Each neuron is identified by the letter Njl, where j describes the neuron number in the layer, and l
is the layer number. Each neuron has an input, the weight of which is given by the value wijl, where
i is the number of the neuron’s input; j describes the neuron number in the layer; l is the layer
number. The neurons in the input layer do not perform any computational operations; they solely

6
DEMO VERSION OF THE SECOND ROUND TASKS

distribute the input signals. OUTjl is the output value of the neuron. (X1, X2) is the input vector,
i.e. the vector that is fed to the input of the network. (Y1, Y2) is the output vector that the neural
network generates as a result of work. For the output layer OUTjl = Yj. The activation functions of
each neuron are sigmoid (logistic function).

The perceptron is trained using the backpropagation algorithm without a training schedule. The
algorithm uses the concept of error gradient for each neuron—∇Ejl . The rms error function is used
as the error.

Weights of neurons in the hidden layer:

W112 W212 W122 W222
0.7599 0.1452 0.9608 0.6935

Output layer neuron weights:

W113 W213 W123 W223
-0.2808 -0.1997 -0.9667 -0.6111

The input vector:

X1 X2
-0.5851 -0.5826

For a given input vector, an output vector is expected (training vector):

D1 D2
0.2030 -0.8542

Assignment:
1. Write the neuron activation formula.
2. Write the formula for calculating the outputs of neuron N12.
3. Write the formula for calculating the error gradient for the output layer neuron N23.
4. Write the formula for calculating the error gradient for the hidden layer neuron N22.
5. Write the formula for calculating the correction of the weights of neuron N13.
6. Calculate the output of each neuron OUTjl.
7. Calculate all error gradients ∇Ejl .
8. Calculate all new weights Wijl.
Round all calculation results to four decimal places (10-4).

Please note that the evaluation will consider your problem-solving process; providing only the
final answer is not sufficient.

Solution:

1. Neuron activation formula.

1
sig(x) =
1 + e−x

7
DEMO VERSION OF THE SECOND ROUND TASKS

2. The formula for calculating the outputs of neuron N12.

OUT12 = sig(x1 ∗ w112 + x2 ∗ w212 )

3. The formula for calculating the error gradient for the output layer neuron N23.
∇E23 = Y2 ∗ (1 − Y2 ) ∗ (Y2 − D2 )

4. The formula for calculating the error gradient for the hidden layer neuron N22.
∇E22 = OUT22 ∗ (1 − OUT22 ) ∗ (∇E13 ∗ w213 + ∇E23 ∗ w223 )

5. The formula for calculating the correction of the weights of neuron N13.
w113 = w113 − ∇E13 ∗ OUT12
w213 = w213 − ∇E13 ∗ OUT22

6. The outputs of each neuron OUTjl.

OUT12 = sig(x1 ∗ w112 + x2 ∗ w212 ) = sig(−0,44461749 − 0,08459352) = 0,3707

OUT22 = sig(x1 ∗ w122 + x2 ∗ w222 ) = sig(−0,56216408 − 0,4040331) = 0,2756

OUT13 = sig(OUT12 ∗ w113 + OUT22 ∗ w213 ) = sig(−0,10409256 − 0,05503732) = 0,4603

OUT23 = sig(OUT12 ∗ w123 + OUT22 ∗ w223 ) = sig(−0,35835569 − 0,16841916) = 0,3713

7. Gradients Ejl.

∇E13 = Y1 ∗ (1 − Y1 ) ∗ (Y1 − D1 ) = OUT13 ∗ (1 − OUT13 ) ∗ (OUT13 − D1 )

= 0,4603 ∗ (1 − 0,4603) ∗ (0,4603 − 0,2030) = 0,0639

∇E23 = Y2 ∗ (1 − Y2 ) ∗ (Y2 − D2 ) = OUT23 ∗ (1 − OUT23 ) ∗ (OUT23 − D2 )

= 0,3713 ∗ (1 − 0,3713) ∗ (0,3713 + 0,8542) = 0,2861

∇E12 = OUT12 ∗ (1 − OUT12 ) ∗ (∇E13 ∗ w113 + ∇E23 ∗ w123 )

= 0,3707 ∗ (1 − 0,3707) ∗ (0,0639 ∗ −0,2808 + 0,2861 ∗ −0,9667)
= −0,0687

∇E22 = OUT22 ∗ (1 − OUT22 ) ∗ (∇E13 ∗ w213 + ∇E23 ∗ w223 )

= 0,2756 ∗ (1 − 0,2756) ∗ (0,0639 ∗ −0,1997 + 0,2861 ∗ −0,6111)
= −0,0375

8. New weights Wijl.

w112 = w112 − ∇E12 ∗ x1 = 0,7599 + 0,0687 ∗ −0,5851 = 0,7197

w212 = w212 − ∇E12 ∗ x2 = 0,1452 + 0,0687 ∗ −0,5826 = 0,1052

w122 = w122 − ∇E22 ∗ x1 = 0,9608 + 0,0375 ∗ −0,5851 = 0,9389

8
DEMO VERSION OF THE SECOND ROUND TASKS

w222 = w222 − ∇E22 ∗ x2 = 0,6935 + 0,0375 ∗ −0,5826 = 0,6717

w113 = w113 − ∇E13 ∗ OUT12 = −0,2808 − 0,0639 ∗ 0,3707 = −0,3045

w213 = w213 − ∇E13 ∗ OUT22 = −0,1997 − 0,0639 ∗ 0,2756 = −0,2173

w123 = w123 − ∇E23 ∗ OUT12 = −0,9667 − 0,2861 ∗ 0,3707 = −1,0727

w223 = w223 − ∇E23 ∗ OUT22 = −0,6111 − 0,2861 ∗ 0,2756 = −0,6899

Answer:
1.
1
sig(x) =
1 + e−x

2.
OUT12 = sig(x1 ∗ w112 + x2 ∗ w212 )

3.
∇E23 = Y2 ∗ (1 − Y2 ) ∗ (Y2 − D2 )

4.
∇E22 = OUT22 ∗ (1 − OUT22 ) ∗ (∇E13 ∗ w213 + ∇E23 ∗ w223 )

5.
w113 = w113 − ∇E13 ∗ OUT12
w213 = w213 − ∇E13 ∗ OUT22

6.
OUT12 OUT 22 OUT 13 OUT 23
0.3707 0.2756 0.4603 0.3713

7.
∇E12 ∇E22 ∇E13 ∇E23
-0.0687 -0.0375 0.0639 0.2861

8.
W112 W212 W122 W222 W113 W213 W123 W223
0.7197 0.1052 0.9389 0.6717 -0.3045 -0.2173 -1.0727 -0.6899

Assessment criteria:
1. Correctly recording the neuron activation formula, the formula for calculating the outputs
of neuron N12, the formula for calculating the gradient for the hidden layer neuron N22, and
the formula for calculating the correction of neuron N13 weights earns 3 points.

9
DEMO VERSION OF THE SECOND ROUND TASKS

2. Calculating the output of each OUTjl. neuron and its error gradient is ∇Ejl earns 6 points.

3. Calculating all new Wijl weights is worth 6 points.

Computer Science: Interdisciplinary Applications

Task 17
Entry level (1 point)

There is a public key cryptosystem with the following parameters: eA is the public key of subscriber
A; eB is the public key of subscriber B; dA is the private key of subscriber A; dB is the private key
of subscriber B. Subscriber A sends a message to subscriber B. With which key does subscriber A
encrypt the message?
a) eA
b) eB
c) dA
d) dB
Answer: b.

Task 18
Entry level (1 point)

Consider an additive Fibonacci generator with the operation equation Xn = (Xn—1 + Xn—2) mod 17.
At a certain point, the generator first produced the number 9, and then the number 11. What number
will appear next in the output?
a) 3
b) 5
c) 10
d) 13
Answer: a.

Task 19
Entry level (1 point)

What is the multiplicative inverse of 3 mod 17?

a) 1
b) 3
c) 6
d) 13
Answer: c.

10
DEMO VERSION OF THE SECOND ROUND TASKS

Task 20
Intermediate level (4 points)

An organization's security officer suspects that there is an insider in the company, communicating
with competitors through a steganographic channel. One of the employees became nervous when
interviewed by the security officer. After intercepting the information flow from that employee's
computer, the security officer realized that the employee in question was trying to send a distress
signal. What is the name of the insider's accomplice?
The intercepted fragment in hexadecimal form is:
F4 F1 F4 FC F4 F9 F4 F3 F4 F5 F2 FC F4 F8 F4 F5 F4 FC F5 F0

Table 1. ASCII Table Fragment

hex char hex char
2C , 4D M
2E . 4E N
41 A 4F O
42 B 50 P
43 C 51 Q
44 D 52 R
45 E 53 S
46 F 54 T
47 G 55 U
48 H 56 V
49 I 57 W
4A J 58 X
4B K 59 Y
4C L 5A Z
Answer: ALICE

Task 21
Advanced level (15 points)

The round function of the block cipher takes a 16-bit value as input. The round function includes
the following operations:
1) Performing an XOR between the original value and the key;
2) Splitting the resulting value into nibbles and replacing each nibble according to Table 2.

Table 2. Round Function Substitution Table

Input 0 1 2 3 4 5 6 7 8 9 A B C D E F
Output 3 E 5 9 A D 8 D F 4 1 0 7 6 B 2

3) Performing a cyclic shift of the obtained value by 3 positions to the left.

What is the value of the lower (right) half of the encrypted message if the original message is
3D7B4AE3h and the upper (left) half of the encrypted message is 35FAh?

11
DEMO VERSION OF THE SECOND ROUND TASKS

Please note that the evaluation will consider your problem-solving process; providing only the
final answer is not sufficient.

Solution:
We begin by decrypting the upper part of the encrypted message.
Let us convert 35FAh to binary form:
35FАh = 0011 0101 1111 1010
and perform a cyclic shift to the right by 3 positions. Thus, we obtain:
0100 0110 1011 1111 = 46BFh
(A correctly performed shift is worth 3 points.)
Then we perform the inverse transformation, according to the table.
46BFh → 9DE8h
(A correctly performed inverse substitution, according to the table, is worth 3 points.)
Then we convert the obtained value and the value of the upper half of the original message into
binary form:
9DE8h = 1001 1101 1110 1000,
37BDh = 0011 0111 1011 1101,
and calculate their sum modulo 2 to obtain the key value.
1001 1101 1110 1000
xor
0011 0111 1011 1101
=
1010 1010 0101 0101
(A correctly inverted XOR operation is worth 3 points.)
(In total, correctly finding the key is worth 3 + 3 + 4 = 9 points.)
Let us begin encrypting the lower half of the original text. First, we convert it into binary form:
4AE3h = 0100 1010 1110 0011
and add modulo 2 with the key

0100 1010 1110 0011

xor
1010 1010 0101 0101
=
1110 0000 1011 0110
(Correctly adding with the key is worth 2 points.)
Then we convert the obtained value into hexadecimal form:
1110 0000 1011 0110 = E0B6h
and perform the substitution according to the table:
E0B6h → B308h = 1011 0011 0000 1000
(Correctly substituting according to the table is worth 2 points.)
All that remains is cyclically shifting the value to the left by 3 positions and obtaining the result
1001 1000 0100 0101 = 9825h
(A correctly performed shift earns 2 points.)

12
DEMO VERSION OF THE SECOND ROUND TASKS

(In total, performing the round function with the known key is worth 2 + 2 + 2 = 6 points)
Answer: 9825h

Computer Science: Software Engineering

Task 22
Entry level (1 point)

In which lifecycle model are all software requirements clearly defined before development begins?
a) Spiral
b) Waterfall
c) Incremental
d) Iterative
Answer: b.

Task 23
Entry level (1 point)

Which of the above is a methodology of “flexible” software development?

a) V-Model
b) Agile
c) Waterfall
d) All of the above
Answer: b.

Task 24
Entry level (1 point)

Which language is classified as declarative?

a) C, C++
b) Assembler, Pascal
c) PHP, Python
d) HTML, SQL
Answer: d.

Task 25
Intermediate level (3 points)

Which TWO stages does the software development life cycle include?
a) Requirements analysis
b) Testing
c) Compilation
d) Discussion

13
DEMO VERSION OF THE SECOND ROUND TASKS

Answer: a, b.

Task 26
Intermediate level (3 points)

Which TWO of the following actions does the user perform during the testing phase?
a) syntax debugging
b) test selection
c) defining the test method
d) determining the form for issuing results
Answer: b, c.

Computer Science: Hardware and Architecture

Task 27
Entry level (1 point)

Which technologies are used to increase the reliability and speed of a hard disk drive?
a) increasing the capacity of the hard disk drive
b) reducing the number of revolutions of the hard disk drive
c) RAID
d) longitudinal recording to disk
Answer: c.

Task 28
Entry level (1 point)

Which of the following statements are true about low-level formatting?

a) Low-level formatting is the layout of a disk into tracks and sectors.
b) Low-level formatting is the process of writing a file system structure to a disk that allows
the disk to be used by the operating system to store programs and data.
c) Low-level formatting should be performed regularly.
d) Low-level formatting depends on the file system type.
Answer: a.

Task 29
Entry level (1 point)

Which of the following statements are true about processor commands?

a) The command currently executed by the processor is located in the controller.
b) The instruction currently being executed by the processor is located in the processor status
register.
c) Any command executed changes the value of the processor status register.

14
DEMO VERSION OF THE SECOND ROUND TASKS

d) A processor instruction has only one operand.

Answer: c.

Task 30
Intermediate level (3 points)

Which of the following statements align with the principles of Von Neumann architecture?
a) A program is written in memory as a sequential code.
b) A processor, having finished executing one instruction, moves on to the next.
c) The address of the next instruction is always the operand of the current instruction.
d) There is RAM for storing instructions and separate RAM for storing data.
e) Memory consists of numbered cells, any of which is accessible to the processor.
f) The binary-decimal number system is used to represent data.
Answer: a, b, e.

Task 31
Intermediate level (3 points)

Which of the following tasks does the Cache Controller perform?

a) providing quick access to intensively used data
b) preloading data into cache from RAM
c) loading data into RAM from the hard drive
d) lazy writing of data from cache memory to RAM
e) loading data into cache memory from a network resource
f) monitoring the execution of instructions by the processor
Answer: a, b, d.

Telecommunications

Task 32
Entry level (1 point)

Which information does the switch collect about connected devices based on the packet
information?
a) The sender's IP address
b) The recipient’s IP address
c) The sender's MAC address
d) The recipient’s MAC address
Answer: c.

Task 33
Entry level (1 point)

15
DEMO VERSION OF THE SECOND ROUND TASKS

Why is it necessary to specify the DNS server address?

a) to determine the MAC address from the IP address
b) to obtain the IP address of the resource
c) to determine the DNS name from the IP address
d) to obtain the DHCP server address
Answer: b.

Task 34
Entry level (1 point)

What is the ping utility used for?

a) To check whether the MAC address matches the IP address
b) To check the download speed of files
c) To test the presence of devices on the network
d) To create data backups
Answer: c.

Task 35
Intermediate level (3 points)

A router is a device that:

a) Connects networks.
b) Creates virtual networks.
c) Connects devices in a network.
d) Routes packets between networks.
Answer: a, d.

16