Chapter 2: Advanced Concepts of Modelling in AI

Section A: Multiple Choice Questions (20 Questions × 1 mark)

1. Supervised learning uses:

a) Unlabelled data b) Reinforcement data c) Labelled data d) No data
2. Unsupervised learning is used to:
a) Predict stock prices b) Cluster similar data c) Calculate taxes d) Identify labeled outputs
3. Reinforcement learning is based on:
a) Clustering b) Reward and punishment c) Input-output pairs d) Supervised data
4. An example of supervised learning is:
a) Email spam detection b) Customer segmentation c) Recommender systems d) Game playing
5. Which is a type of supervised learning problem?
a) Classification b) Clustering c) Association d) Filtering
6. Which model is best suited for predicting housing prices?
a) Classification b) Regression c) Clustering d) Association
7. Rule-based systems are:
a) Hand-coded by humans c) Learned by machines
b) Always neural networks d) Based on reinforcement
8. The first AI models were:
a) Rule-based b) Clustering models
b) Association models d) Reinforcement learners
9. A neural network is made up of:
a) Only inputs b) Neurons and layers c) Labels only d) Rules only
10. The layer that performs computation in a neural network is:
a) Input layer b) Output layer c) Hidden layer d) Feedback layer
11. A perceptron is:
a) A clustering algorithm b) An output classifier c) A basic neural unit d) A dataset
12. Deep learning is:
a) Rule-based AI c) Always supervised
b) Multi-layered neural networks d) Unsupervised learning
13. The main function of the output layer is:
a) Receiving input b) Generating results c) Performing calculations d) Adjusting weights
14. Clustering is mainly used in:
a) Classification b) Unsupervised learning c) Regression d) Rule-based models
15. Which of the following is a reinforcement learning application?
a) Self-driving cars b) Weather forecasting c) Sorting emails d) Image tagging
16. Which AI model works without labeled data?
a) Supervised learning b) Reinforcement learning c) Unsupervised learning d) Rule-based AI
17. In supervised learning, what do we provide the machine?
a) Rules b) Input-output pairs c) Hidden patterns d) Rewards
18. Which learning method uses a reward system?
a) Reinforcement learning b) Supervised learning c) Unsupervised learning d) Regression
19. Neural networks learn by:
a) Rule following b) Adjusting weights c) Clustering d) Output estimation
20. Which model learns from experience like a human?
a) Rule-based model b) Reinforcement model c) Clustering model d) Association model

Section B: Very Short Answer Questions (20 Questions × 1 Mark)

1. What is supervised learning?

2. What is clustering?
3. What does a neural network mimic?
4. Name one application of reinforcement learning.
 5. What is the purpose of the input layer?
6. Define a perceptron.
7. What is classification?
8. What kind of data is used in unsupervised learning?
9. What is the full form of ANN?
10. What does the hidden layer do?
11. What is an output layer?
12. What is deep learning?
13. Give one example of supervised learning.
14. What does reinforcement learning use for learning?
15. Which AI model uses trial and error?
16. What is the main task of regression?
17. Define rule-based AI.
18. Give an example of unsupervised learning.
19. What is the use of association models?
20. What is a feature in a dataset?

Section C: Assertion-Reason Questions (Based on CBSE Format)

Direction: For the following questions, choose the correct option:
• (A) Both Assertion (A) and Reason (R) are true, and R is the correct explanation of A.
• (B) Both Assertion (A) and Reason (R) are true, but R is not the correct explanation of A.
• (C) Assertion (A) is true, but Reason (R) is false.
• (D) Assertion (A) is false, but Reason (R) is true.

1. Assertion (A): Supervised learning requires labeled data.

Reason (R): In supervised learning, the output is unknown for the input data.

2. Assertion (A): Clustering is used in unsupervised learning.

Reason (R): Clustering groups similar data points without prior labels.

3. Assertion (A): Reinforcement learning agents learn by receiving feedback from their actions.
Reason (R): Reinforcement learning provides rewards or penalties based on agent behavior.

4. Assertion (A): Rule-based AI is dynamic and learns from data.

Reason (R): Rule-based systems adapt over time using past experiences.

5. Assertion (A): Neural networks are inspired by the structure of the human brain.
Reason (R): Neurons in neural networks work exactly like biological neurons.

6. Assertion (A): Deep learning can automatically extract features from raw data.
Reason (R): Deep learning models consist of multiple hidden layers.

7. Assertion (A): Association rule learning is a type of supervised learning.

Reason (R): It requires labeled outputs to work.

8. Assertion (A): Regression models produce continuous output values.

Reason (R): Regression is a type of supervised learning.

9. Assertion (A): A perceptron is the most complex neural unit.

Reason (R): It is made up of many interconnected layers.

10. Assertion (A): Classification problems deal with discrete outputs.

Reason (R): Email spam detection is a classification problem.
11. Assertion (A): Reinforcement learning needs labeled input-output pairs.
Reason (R): It is a type of supervised learning.

12. Assertion (A): Deep neural networks may require more computational resources than simple ML models.
Reason (R): They have more layers and parameters to train.

13. Assertion (A): Unsupervised learning is used when no training data is available.
Reason (R): The model is expected to learn structure from data without output labels.

14. Assertion (A): The hidden layer in a neural network is responsible for displaying final output.
Reason (R): Output layer performs all internal computations.

15. Assertion (A): Association models can be used to recommend products.

Reason (R): These models identify relationships between items.

Section D: Short Answer Questions (2 Marks each)

1. Differentiate between supervised and unsupervised learning.
2. What is reinforcement learning? Give an example.
3. Explain classification and give one example.
4. Define regression. Mention one use case.
5. What is a neural network?
6. What does a perceptron consist of?
7. Explain rule-based AI with an example.
8. Why are hidden layers important?
9. Define deep learning. How is it different from ML?
10. What is clustering used for?
11. What are labeled and unlabeled data?
12. What is meant by the learning rate in neural networks?
13. Give an example of classification and regression.
14. How does reinforcement learning improve decision-making?
15. Explain the structure of a neural network.
16. Why is unsupervised learning used?
17. What are the challenges of supervised learning?
18. Mention any two real-life applications of ANN.
19. What does the output layer do in ANN?
20. What is association rule mining?

Section E: Long Answer Questions (4 Marks each)

1. Explain the three types of AI learning models with examples.
2. Describe the structure and function of Artificial Neural Networks.
3. Compare supervised, unsupervised, and reinforcement learning in tabular form.
4. What is deep learning? Discuss its advantages and limitations.
5. Explain how a perceptron works with a diagram.
6. Describe rule-based vs. learning-based models with examples.
7. Illustrate the flow of a neural network with layers and their roles.
8. Explain classification and regression in detail with applications.
9. Write the differences between ANN and CNN.
10. Discuss reinforcement learning and its working using a reward system.