MODEL TEST PAPER (XII AI)

MODEL TEST PAPER – 3

Class – XII                                                                              Subject – Artificial Intelligence
Max. Time: 2 Hours                                                              Max. Marks: 50

---

General Instructions:

1. This Question Paper consists of 21 questions divided into two sections: Section A and Section B.
2. Section A contains Objective Type Questions (24 marks), while Section B consists of Subjective Type Questions (26 marks).
3. Out of the given 21 questions, you have to answer 15 questions (5 from Employability Skills and 10 from Subject Specific Skills).
4. All questions in a section must be answered in the correct sequence.

---

SECTION A: OBJECTIVE TYPE QUESTIONS            (24 Marks)

1. Answer any 4 out of the given 6 questions on Employability Skills.     (1 x 4 = 4)

1. What does the ‘M’ in SMART goals stand for?

a) Measurable

b) Manageable

c) Marketable

d) Motivational

2. Communication skills are enhanced by:

a) Speaking louder

b) Interrupting others

c) Active listening and feedback

d) Ignoring non-verbal cues

3. Which of the following is an example of self-management?

a) Managing others’ schedules

b) Organizing your own tasks effectively

c) Delegating work to others

d) Avoiding responsibilities

4. Entrepreneurial traits include:

a) Risk-taking

b) Passive decision-making

c) Avoiding innovation

d) Delegating all tasks

5. True or False: Non-verbal communication involves only spoken words.
6. Time management is most effective when:

a) Deadlines are ignored

b) Goals are unclear

c) Tasks are prioritized

d) Time is wasted on unimportant activities

---

2. Answer any 5 out of the given 6 questions.         (1 x 5 = 5 marks)

1. Supervised Learning is used when:

a) Data is labeled

b) Data is unlabeled

c) Clustering is required

d) No data is available

2. Which of the following is a method for evaluating model performance?

a) Cross-validation

b) Regularization

c) Data augmentation

d) None of the above

3. The AI Project Cycle includes which of the following phases?

a) Problem Definition

b) Data Gathering

c) Model Evaluation

d) All of the above

4. The Mean Squared Error (MSE) is calculated by:

a) Taking the square root of the difference between predicted and actual values

b) Taking the average of the squared differences between predicted and actual values

c) Summing all predicted values

d) Subtracting the predicted values from actual values

5. Data visualization helps in:

a) Hiding important data points

b) Presenting data in a clear and insightful manner

c) Making data less interpretable

d) Reducing the accuracy of data analysis

6. Normalization is used to:

a) Reduce data size

b) Scale features to a similar range

c) Eliminate outliers

d) Improve model complexity

---

3. Answer any 5 out of the given 6 questions.         (1 x 5 = 5 marks)

1. Overfitting occurs when:

a) The model performs well on both training and test data

b) The model performs well on training data but poorly on test data

c) The model performs poorly on all datasets

d) The model performs equally well on training and test data

2. Hyperparameters are:

a) Parameters that can be changed by the model during training

b) Parameters set before the training starts

c) Internal parameters adjusted by the model

d) Values assigned randomly

3. True or False: Data gathering is the final stage of the AI Project Cycle.
4. Precision and recall are metrics used to evaluate:

a) Regression models

b) Classification models

c) Clustering models

d) Time-series models

5. The K-nearest neighbors (KNN) algorithm is used for:

a) Regression

b) Classification

c) Clustering

d) All of the above

6. Principal Component Analysis (PCA) is used for:

a) Dimensionality reduction

b) Data scaling

c) Data normalization

d) Feature selection

---

4. Answer any 5 out of the given 6 questions.         (1 x 5 = 5 marks)

1. Cross-validation is mainly used to:

a) Reduce overfitting

b) Reduce underfitting

c) Improve model complexity

d) Reduce training time

2. True or False: Data visualization is not necessary for data storytelling.
3. Clustering is used in unsupervised learning to:

a) Classify labeled data

b) Group similar data points

c) Predict continuous outcomes

d) Identify patterns in structured data

4. Confusion matrix is used to measure:

a) Regression accuracy

b) Classification accuracy

c) Clustering effectiveness

d) Time-series forecasting

5. The process of feature engineering helps in:

a) Improving the performance of models

b) Reducing data noise

c) Optimizing data collection

d) Increasing data size

6. True or False: Deep learning models perform well on small datasets with limited features.

---

5. Answer any 5 out of the given 6 questions.                     (1 x 5 = 5 marks)

1. Mean Absolute Percentage Error (MAPE) is used for:

a) Measuring accuracy in classification models

b) Measuring error in regression models

c) Calculating precision

d) Evaluating clustering results

2. Supervised Learning can be used in:

a) Classification problems

b) Regression problems

c) Both a & b

d) None of the above

3. Decision trees are used for:

a) Regression

b) Classification

c) Both a & b

d) Clustering

4. True or False: Reinforcement Learning is based on reward and punishment mechanisms.
5. Data cleaning is a crucial step to:

a) Remove irrelevant data

b) Eliminate missing values

c) Improve data quality

d) All of the above

6. Cross-entropy loss is used in:

a) Regression models

b) Classification models

c) Time-series models

d) Clustering models

---

SECTION B: SUBJECTIVE TYPE QUESTIONS                      (26 Marks)

6. Answer any 3 out of the given 5 questions on Employability Skills.     (2 x 3 = 6)

1. What is the significance of goal setting in entrepreneurship?
2. Explain the role of time management in achieving personal and professional success.
3. How does non-verbal communication enhance interactions?
4. Describe the difference between intrinsic and extrinsic motivation.
5. What are the key traits of a successful entrepreneur?

---

7. Answer any 4 out of the given 6 questions.         (2 x 4 = 8 marks)

1. What are the key phases in the AI Project Life Cycle?
2. Explain the importance of cross-validation in machine learning.
3. Describe the Mean Absolute Error (MAE) and its importance in evaluating model performance.
4. How does data storytelling influence decision-making in AI?
5. What are the different types of learning algorithms in AI, and where are they applied?
6. Explain how problem scoping is done in AI project development.

---

8. Answer any 3 out of the given 5 questions.         (4 x 3 = 12 marks)

1. Explain the concept of overfitting and how it can be prevented.
2. Describe the steps involved in building an AI model from data collection to deployment.
3. Explain the differences between supervised, unsupervised, and reinforcement learning.
4. Using the following data, calculate the RMSE:
   Predicted: [15, 25, 35, 45]
   Actual: [14, 24, 33, 46]
5. What are the ethical considerations in AI, and why are they important for responsible AI development?

---

MARKING SCHEME

SECTION A: OBJECTIVE TYPE QUESTIONS                        (24 Marks)

---

1. Answer any 4 out of the given 6 questions on Employability Skills.

1. a) Measurable
2. c) Active listening and feedback
3. b) Organizing your own tasks effectively
4. a) Risk-taking
5. False
6. c) Tasks are prioritized

---

2. Answer any 5 out of the given 6 questions.

1. a) Data is labeled
2. a) Cross-validation
3. d) All of the above
4. b) Taking the average of the squared differences between predicted and actual values
5. b) Presenting data in a clear and insightful manner
6. b) Scale features to a similar range

---

3. Answer any 5 out of the given 6 questions.

1. b) The model performs well on training data but poorly on test data
2. b) Parameters set before the training starts
3. False
4. b) Classification models
5. b) Classification
6. a) Dimensionality reduction

---

4. Answer any 5 out of the given 6 questions.

1. a) Reduce overfitting
2. False
3. b) Group similar data points
4. b) Classification accuracy
5. a) Improving the performance of models
6. False

---

5. Answer any 5 out of the given 6 questions.

1. b) Measuring error in regression models
2. c) Both a & b
3. c) Both a & b
4. True
5. d) All of the above
6. b) Classification models

---

SECTION B: SUBJECTIVE TYPE QUESTIONS                      (26 Marks)

---

6. Answer any 3 out of the given 5 questions on Employability Skills.

1. Goal setting helps entrepreneurs define clear objectives and focus their efforts on achieving them. It gives direction and purpose, ensuring that resources are allocated efficiently. Without proper goals, entrepreneurs may lack focus and direction in their business ventures.
2. Time management is crucial for achieving both personal and professional success. It involves organizing tasks, setting priorities, and adhering to deadlines. Effective time management reduces stress and allows individuals to accomplish more in less time.
3. Non-verbal communication enhances interactions by conveying emotions and intentions beyond spoken words. Facial expressions, gestures, and posture can emphasize key points, demonstrate attentiveness, and help build rapport.
4. Intrinsic motivation comes from within and is driven by personal satisfaction and interest in the task itself. Extrinsic motivation, on the other hand, is driven by external rewards like money or recognition. Intrinsic motivation often leads to higher job satisfaction and better performance.
5. Successful entrepreneurs possess key traits such as risk-taking, innovation, resilience, and vision. These traits help them navigate the uncertainties of business, adapt to change, and capitalize on opportunities.

---

7. Answer any 4 out of the given 6 questions.

1. The AI Project Life Cycle includes the following phases:
   • Problem Scoping: Identifying the problem and understanding its scope.
   • Data Collection: Gathering relevant data to build the AI model.
   • Data Preparation: Cleaning and preparing data for analysis.
   • Model Building: Selecting algorithms and training the model.
   • Model Evaluation: Assessing the model’s performance.
   • Model Deployment: Deploying the model for real-world use.
2. Cross-validation helps evaluate how well a model generalizes to new data. It is particularly useful in preventing overfitting, ensuring the model doesn’t simply memorize the training data. K-Fold cross-validation is one of the most commonly used techniques.
3. Mean Absolute Error (MAE) measures the average magnitude of prediction errors in a regression model. It is calculated by averaging the absolute differences between predicted and actual values. Lower MAE values indicate better model performance.
4. Data storytelling makes data insights more accessible by turning complex data findings into compelling narratives. It helps decision-makers understand the key takeaways from data and make informed decisions. Visual aids like graphs and charts are often used to complement the story.
5. The three types of learning algorithms are:
   • Supervised Learning: Uses labeled data for training (e.g., classification).
   • Unsupervised Learning: Works with unlabeled data to find hidden patterns (e.g., clustering).
   • Reinforcement Learning: Learns by interacting with the environment and receiving feedback (e.g., game playing).
6. Problem scoping in AI projects involves defining the problem clearly, identifying the stakeholders, understanding the constraints, and setting success criteria. This phase ensures the AI solution aligns with business objectives and user needs.

---

8. Answer any 3 out of the given 5 questions.

1. Overfitting occurs when a model performs very well on training data but poorly on new, unseen data. It can be prevented by using techniques such as cross-validation, regularization, and reducing the complexity of the model (e.g., pruning decision trees).
2. Steps in building an AI model:
   • Data Collection: Acquire raw data.
   • Data Cleaning: Remove noise, handle missing data.
   • Feature Engineering: Select/create useful input features.
   • Model Training: Use algorithms to build the model.
   • Model Evaluation: Test the model using a validation set.
   • Deployment: Implement the model into real-world applications.
3. Supervised learning uses labeled data to train models for tasks like classification and regression. Unsupervised learning works with unlabeled data to find hidden patterns, while reinforcement learning relies on reward feedback to improve the model’s performance over time.
4. RMSE Calculation for the data:
   Predicted: [15, 25, 35, 45]
   Actual: [14, 24, 33, 46]
   RMSE = √[(1² + 1² + 2² + 1²) / 4]
   RMSE = √[(1 + 1 + 4 + 1) / 4] = √7 / 4 = √1.75 ≈ 1.32
5. Ethical considerations in AI involve ensuring fairness, transparency, and accountability in AI systems. AI models must avoid biases, protect user privacy, and be transparent about how decisions are made. These considerations are crucial for building trust and ensuring responsible AI development.