DEPARTMENT OF MATHEMATICAL & PHYSICAL SCIENCES
Chapter wise Question Set
DSA 5011 (AI and Deep Learning), Fall 2024
Chapter- 1: Introduction
1. What do you understand by Artificial Intelligence (AI)? Give some real world problems
to solve using AI techniques.
2. Why study AI? What are the areas of AI? Explain.
3. Write down the advantages and disadvantages of Artificial Intelligence (AI).
4. List and explain some applications of AI.
5. Define Intelligence. Make a relationship between Brain Vs Computer.
6. Write down the characteristics/properties of Artificial Intelligence (AI) problem.
7. How difficult is it to achieve AI?
8. How do you differentiate between strong and weak AI?
Chapter-2: Logical/Intelligent Agents
1. Give a PEAS description of the task environment for the following agents:
(i) Vacuum-cleaner world
(ii) An intelligent agent that solves Rubik’s Cube.
(iii) A drone to deliver commercial packages in a city.
(iv) Wumpus World.
Figure: 3×3 Rubik’s Cube
2. For each of the above scenarios (Question no. 1), characterize and discuss the
environment according to whether it is fully or partially observable, static or dynamic,
deterministic or stochastic, episodic or sequential, discrete or continuous.
3. What is the expansion if PEAS in task environment?
4. Write short notes on (i) Human Agents and (ii) Robotic Agents.
5. Make a relationship between Agents and Environment. Give Example(s).
6. What are the types of intelligent Agents? Explain each of them shortly.
7. Write down the differences between Single Agents and Multi Agents.
8. What is Turing Test? Explain with example(s).
9. Make a relationship between static and dynamic agent.
Chapter-2: Neural Networks
1. Make a relationship between Biological Neural Network (BNN) and Artificial Neural Network
(ANN).
2. What do you mean by Neural Networks with suitable example(s).
1. What is neuron? Describe the structure of a neuron and explain how it operates?
2. Write down the applications of Neural Networks.
3. What does a neural network consist of? How does a neural network learn things?
4. What is back-Propagation in Neural Network?
5. What is the ability of a single neuron with a non-linear activation function? Can it only classify
the input space in two classes?
6. What is bias in artificial neural network?
7. How to solve XOR problem through neural networks (NNs)
8. Activation functions in Neural Networks? What is the role of Activation function of Neural
Network?
9. Exercise: Consider the following Neural Network (NN), Update each weight and then calculate
the output. In this case, use the Sigmoid as the activation function.
10. Consider the following recurrent neural network, in which the bottom node is the input, the top
node is the output, and the middle two nodes are hidden.
(i) Draw an equivalent, feed-forward (non-recurrent) neural network representing three time
instances.
(ii) Add weights to your feed-forward network, clearly indicating which weights are equal.
(iii) Mark the corresponding weights on the recurrent network above.
Chapter-3: Deep Learning-Recurrent Neural Network (RNN)
1. Define Deep Learning. Give the Example of Deep Learning.
2. How the task could be addressed/solved by a many-to-one Recurrent Neural Network
(RNN) model architecture.
(i) Movie review (positive/negative) classification,
(ii) Gender recognition from audio.
(iii) Character level language recognition.
3. Define recurrent neural network (RNN). Which type of problems solve by RNN. Give
example(s).
4. Why Recurrent Neural Networks (RNNs) Dominate Sequential Data Analysis?
5. Write down the steps for training Recurrent Neural Networks (RNNs).
6. How Recurrent Neural Networks (RNNs) differs from feedforward Neural Network.
7. What are the types of Recurrent Neural Networks (RNNs). Explain each of them shortly
with example(s).
8. Write down the applications of Recurrent Neural Networks (RNNs).
9. Draw and explain the architecture of Recurrent Neural Networks (RNNs).
Chapter-4: Deep Learning-Convolutional Neural Network (CNN)
1. What do you mean by Convolutional Neural Network (CNN)?
2. Why do we prefer Convolutional Neural networks (CNN) over Artificial Neural networks
(ANN) for image data as input?
3. Explain the different layers in Convolutional Neural Network (CNN).
4. Explain the significance of the RELU Activation function in Convolution Neural
Network (CNN).
5. Why do we use a Pooling Layer in a CNN? Explain with Example(s).
6. What is the size of the feature map for a given input size image, Filter Size, Stride, and
Padding amount?
7. An input image has been converted into a matrix of size 12 X 12 along with a filter of
size 3 X 3 with a Stride of 1. Determine the size of the convoluted matrix.
8. Briefly explain the two major steps of CNN i.e, Feature Learning and Classification.
9. What are the different types of Pooling? Explain their characteristics.
10. Does the size of the feature map always reduce upon applying the filters? Explain why or
why not.
11. What is the role of the Fully Connected (FC) Layer in CNN?
12. Briefly explain the two major steps of CNN i.e, Feature Learning and Classification.
13. Explain the role of the Convolution Layer in CNN.
14. Can we use CNN to perform Dimensionality Reduction? If Yes then which layer is
responsible for dimensionality reduction particularly in CNN?
***Not Include at Midterm
Chapter-5: Artificial Intelligence Searching Techniques (CNN)
1. Write the sequence of nodes in the order visited by the specified algorithms below.
Assumptions:
The algorithms do not generate paths with loops.
Nodes are selected in the alphabetical order when a tie occurs.
A node is visited when that it is at the front of the search queue.
(i) Depth-first search algorithm.
(ii) Breadth-first search algorithm.
(iii) A* search algorithm. Please also show the total cost and details of the steps.
(iv) Please discuss which search algorithm do you prefer to use for this problem and why?
2. Given the figure below, suppose you want to go from start location “Lugoj” todestination
location“Bucharest”. The route distances are labeled on the edges between towns. The
table on the right gives the straight-line distances from the cities to Bucharest. Write
down the order in which the states are visited, and the path found by the following search
algorithms. Show the steps that the algorithm takes to find the path.
Note: Assume that algorithms execute the goal check when nodes are visited. (i) Greedy
Search (ii) Uniform Cost search algorithm (iii) A* search algorithm. Please also show the
total cost and details of the steps.