National Artificial Olympiad - June 1, 2024

Question Coverage

Topics as per syllabus Coverage

Programming 30%

Machine Learning 15%

Deep Learning 15%

NLP 10%

Transformer 10%

Generative Modeling 10%

Computer Vision 10%

MCQ Samples

There will be 30 MCQs covering all the topic of syllabus

Q: Emily is writing a Python script to automate her data processing tasks. She needs
to read a large CSV file into a DataFrame for analysis. Which library should she use?
A. NumPy
B. SciPy
C. Pandas (Correct)
D. Matplotlib

Q: Olivia is debugging her Python code and needs to print the type of a variable to
the console. Which function should she use?
A. type() (Correct)
B. isinstance()
C. id()
D. print()

Q: You want to build a model to predict stock prices. What's the ideal scenario
regarding bias and variance?
A. High Bias, High Variance
B. Low Bias, Low Variance (Correct)
C. High Bias, Low Variance
D. Low Bias, High Variance
Q: Imagine you're using gradient descent to find the pizza place with the highest
average rating based on user reviews. What's the equivalent of the "cost function" in
this scenario?
A. Number of User Reviews
B. Average Distance from Your Location
C. Total Price of Pizzas
D. Average User Rating (Correct)

Q: You're training a model to classify emails as spam or not spam using gradient
descent. The cost function isn't decreasing after each iteration. What could be the
problem?

A. The data is not rescaled

B. The learning rate is too high
C. The learning rate is too low (Correct)
D. There is not enough data

Q: What is the model representation used by KNN ?

A. set of linear equations
B. A decision tree
C. The entire training dataset (Correct)
D. A neural network architecture

Q: What is the purpose of using a minimum instance count for stopping criteria in
CART ?
A. To ensure all leaves have the same number of data points
B. To prevent the tree from becoming too complex and overfitting the data
(Correct)
C. To improve the accuracy of predictions on unseen data
D. To reduce the computational cost of training the tree

Q: You are working on a binary classification project of classifying if the given text is
spam or not. What activation function will you imply in the last layer of the Dense
layer?
A. ReLU
B. Tanh
C. Sigmoid (Correct)
D. Linear

Q: What is a Generative Adversarial Network (GAN)?

A. A network for data analysis
B. A type of neural network architecture for generative modeling (Correct)
C. A network for fast data processing
D. A network for data encryption
Subjective Questions

There will be 9 Subjective Questions covering all the topics of the syllabus.

Sample Questions are as follows.

Q: Complete the following code so that we can print the missing values using the
‘is_null()’ function.
print("\nMissing Values:")

print( ...................................... )

# Display class distribution of the species feature.

print("\nClass Distribution:")
print(iris_df['species'].value_counts())

Answer:
print(iris_df.is_null().sum())

Q: You are given a list of dictionaries representing employees in a company. Each

dictionary contains the employee's name and their salary. You need to create a new
list that contains the names of all employees who earn more than $50,000. Use a list
comprehension to accomplish this task.

# List of dictionaries representing employees

employees = [ {'name': 'John Doe', 'salary': 60000},

{'name': 'Jane Smith', 'salary': 48000},
{'name': 'Emily Davis', 'salary': 52000},
{'name': 'Michael Brown', 'salary': 45000},
{'name': 'Jessica Taylor', 'salary': 70000}
]

# List comprehension to get names of employees earning more than $50,000

high_earners = .........

# Output the result

print(............................)

Answer:
high_earners = [employee['name'] for employee in employees if
employee['salary'] > 50000]
print(high_earners)
Q: Write the Output of the code and explain.
# List of product names
product_names = ['Laptop', 'Smartphone', 'Tablet', 'Headphones',
'Smartwatch']

# List of corresponding product prices

product_prices = [1000, 800, 300, 150, 200]

# List comprehension to merge the two lists into a list of dictionaries

products = [{'name': name, 'price': price} for name, price in
zip(product_names, product_prices)]

# Output the result

print(products)

Output:

Explanation:

Answer:
[ {'name': 'Laptop', 'price': 1000}, {'name': 'Smartphone',
'price': 800}, {'name': 'Tablet', 'price': 300}, {'name':
'Headphones', 'price': 150}, {'name': 'Smartwatch', 'price': 200}]

Q: You are given two lists, one containing the names of students and the other
containing their corresponding grades. Write a Python function that uses the zip
method to create a dictionary where the student names are the keys and their
grades are the values.

def create_student_grade(names, grades):

# write code below

# List of student names

names = ['Alice', 'Bob', 'Charlie', 'David', 'Eve']

# List of corresponding student grades

grades = ['A', 'B', 'C', 'B+', 'A-']

student_grade_dictionary = create_student_grade(names, grades)

# Output the result

 print(student_grade_dictionary)

Output:

Answer:
def create_student_grade(names, grades):
# write code below
student_grade_dict = dict(zip(names, grades)) \
return student_grade_dict

Q: You are given a 2D NumPy array representing a matrix of integers. Write a Python
script to flatten this matrix into a 1D array. Explain a use case where flattening a
matrix would be useful.

Code:
import numpy as np
# write your code below

Output:

Answer:
matrix = np.array([ [1, 2, 3],
[4, 5, 6],
[7, 8, 9]]
)
# Flatten the matrix into a 1D array
flattened_array = matrix.flatten()

# Output the result

print(flattened_array)