MODEL TEST PAPER (XII AI)

MODEL TEST PAPER – 2

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 marks)

1. What is meant by non-verbal communication?

a) Communication through words

b) Communication through gestures

c) Communication through writing

d) Communication through images

2. Which of the following is a feature of self-motivation?

a) Dependence on others

b) Finding inspiration from external rewards

c) Internal drive to complete tasks

d) Following instructions blindly

3. The ALT + TAB shortcut is used for:

a) Closing a window

b) Switching between windows

c) Refreshing the page

d) Opening the task manager

4. Entrepreneurial competencies include:

a) Delegating tasks only

b) Risk-taking and innovation

c) Avoiding decision-making

d) Managing without any teams

5. A good goal-setting framework focuses on goals that are:

a) Vague and unrealistic

b) SMART (Specific, Measurable, Achievable, Relevant, Time-bound)

c) Only financially oriented

d) Externally driven

6. Active listening requires:

a) Interrupting the speaker often

b) Responding without listening

c) Focused attention and feedback

d) Ignoring non-verbal cues

---

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

1. Which of the following describes a decision tree model?

a) Linear model

b) Probabilistic model

c) Hierarchical model

d) Geometric model

2. Feature selection is done to:

a) Improve model accuracy

b) Reduce overfitting

c) Improve training time

d) All of the above

3. Cross-validation is used to:

a) Increase model accuracy

b) Measure model generalization

c) Reduce overfitting

d) All of the above

4. The Mean Absolute Error (MAE) is used to:

a) Measure model precision

b) Measure model error

c) Improve model complexity

d) Determine model accuracy

5. Supervised Learning requires:

a) No labeled data

b) Labeled training data

c) Clustering techniques

d) Unlabeled data

6. Normalization of data helps to:

a) Scale features to a uniform range

b) Increase model complexity

c) Remove outliers

d) Improve data quality

---

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

1. Which phase involves data cleaning and feature engineering?

a) Model Validation

b) Model Building

c) Data Preparation

d) Model Deployment

2. True or False: Deep learning requires structured data to perform well.
3. What is the role of a validation set in machine learning?

a) To improve training accuracy

b) To evaluate the model performance

c) To reduce dataset size

d) To modify model features

4. Hyperparameter tuning involves adjusting:

a) The model’s internal weights

b) The model’s structural parameters

c) The output layer

d) The input dataset

5. Unsupervised learning is used when:

a) The data has no labels

b) The data is complete

c) The model has overfitted

d) There is more training data available

6. True or False: Gradient Descent is used to minimize the loss function in machine learning models.

---

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

1. Which of the following methods can be used to handle missing data?

a) Imputation

b) Deleting rows

c) Both a & b

d) None of the above

2. Regularization is used in machine learning to:

a) Prevent overfitting

b) Reduce bias

c) Increase the dataset size

d) Improve model accuracy

3. Confusion matrix is used to evaluate:

a) Regression models

b) Classification models

c) Clustering models

d) Data normalization

4. True or False: The AI Project Cycle includes Problem Scoping, Data Gathering, and Model Validation.
5. Which of the following helps in dimensionality reduction?

a) Principal Component Analysis (PCA)

b) Gradient Descent

c) Cross-validation

d) Data augmentation

6. Which of the following is a classification algorithm?

a) K-Nearest Neighbors (KNN)

b) K-Means Clustering

c) Principal Component Analysis (PCA)

d) Support Vector Machines (SVM)

---

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

1. Data storytelling helps in:

a) Presenting insights in an engaging way

b) Making data visualization complex

c) Removing patterns from data

d) Predicting future trends

2. Which of the following is a hyperparameter in machine learning models?

a) Learning rate

b) Model weights

c) Input features

d) Dataset size

3. Overfitting occurs when:

a) The model performs well on training data but poorly on new data

b) The model performs well on all data

c) The model performs poorly on training data

d) The model is too simple

4. True or False: Ensemble learning involves combining multiple models to improve performance.
5. Which of the following techniques is used in unsupervised learning?

a) Regression

b) Classification

c) Clustering

d) Cross-validation

6. Cross-validation helps to:

a) Reduce overfitting

b) Increase training time

c) Measure generalization error

d) Both a & c

---

SECTION B: SUBJECTIVE TYPE QUESTIONS                                  (26 Marks)

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

1. Define active listening and its importance in effective communication.
2. How does self-management impact one’s productivity?
3. Explain the role of non-verbal communication in conveying emotions.
4. Describe the steps involved in goal-setting.
5. How does self-motivation contribute to entrepreneurial success?

---

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

1. What are the main components of the AI Project Cycle?
2. Explain the significance of cross-validation in model evaluation.
3. Describe the use of Mean Absolute Error (MAE) in measuring model performance.
4. How does data visualization enhance storytelling in AI projects?
5. What are some methods of hyperparameter tuning?
6. Describe the process of problem scoping in an AI project.

---

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

1. Explain the role of K-Fold Cross Validation in improving model performance.
2. Describe the steps involved in creating a machine learning model from raw data.
3. What is the difference between classification and regression algorithms? Provide examples.
4. Calculate the RMSE for the following data:
   Predicted: [10, 20, 30, 40]
   Actual: [12, 18, 29, 39]
5. How does EDA (Exploratory Data Analysis) contribute to AI model building?

---

MARKING SCHEME

SECTION A: OBJECTIVE TYPE QUESTIONS                                                (24 Marks)

---

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

1. b) Communication through gestures
2. c) Internal drive to complete tasks
3. b) Switching between windows
4. b) Risk-taking and innovation
5. b) SMART (Specific, Measurable, Achievable, Relevant, Time-bound)
6. c) Focused attention and feedback

---

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

1. c) Hierarchical model
2. d) All of the above
3. d) All of the above
4. b) Measure model error
5. b) Labeled training data
6. a) Scale features to a similar range

---

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

1. c) Data Preparation
2. False
3. b) To evaluate the model performance
4. b) The model’s structural parameters
5. a) The data has no labels
6. True

---

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

1. c) Both a & b
2. a) Prevent overfitting
3. b) Classification models
4. True
5. a) Principal Component Analysis (PCA)
6. d) Support Vector Machines (SVM)

---

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

1. a) Presenting insights in an engaging way
2. a) Learning rate
3. a) The model performs well on training data but poorly on new data
4. True
5. c) Clustering
6. d) Both a & c

---

SECTION B: SUBJECTIVE TYPE QUESTIONS                                  (26 Marks)

---

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

1. Active listening is a communication skill that involves fully concentrating, understanding, and responding thoughtfully to what the speaker is saying. It involves providing feedback, maintaining eye contact, and avoiding interruptions. Active listening is crucial in communication as it fosters trust, ensures clear understanding, and reduces the chances of miscommunication.
2. Self-management is the ability to regulate one’s emotions, thoughts, and behaviors effectively in different situations. It involves goal setting, time management, and maintaining a work-life balance. By prioritizing tasks and staying organized, self-management helps individuals be more productive and achieve their personal and professional goals.
3. Non-verbal communication refers to the transmission of messages without the use of words. It includes gestures, facial expressions, posture, eye contact, and body language. Non-verbal cues can enhance verbal communication by providing additional context and emotional tone, helping to clarify the message.
4. Goal-setting is a critical skill that helps individuals and teams set clear, achievable targets. The SMART framework (Specific, Measurable, Achievable, Relevant, Time-bound) is often used to ensure goals are well-defined and attainable. Clear goals provide direction, motivate performance, and help measure progress.
5. Self-motivation refers to the internal drive to pursue goals and overcome obstacles without needing external encouragement. Traits such as persistence and optimism are essential for achieving success. Entrepreneurs, for example, rely heavily on self-motivation to continue innovating and taking risks, despite challenges.

---

7. Answer any 4 out of the given 6 questions.                     (4*2)

1. The AI Project Cycle consists of several phases:
   • Problem Scoping: Understanding and defining the problem to be solved with AI.
   • Data Collection: Gathering relevant data that will be used for building the model.
   • Data Preparation: Cleaning and organizing the data to make it ready for analysis.
   • Model Building: Choosing and applying algorithms to train the model.
   • Model Evaluation: Testing the model to assess its accuracy and performance.
   • Model Deployment: Implementing the model into real-world applications.
2. Cross-validation is a statistical method used to evaluate the performance of machine learning models. In K-Fold cross-validation, the dataset is split into K smaller sets. The model is trained on K-1 of these sets and tested on the remaining set. This process repeats K times, with each set used once as the test set. It helps reduce overfitting and provides a more accurate measure of model performance.
3. Mean Absolute Error (MAE) is a metric that measures the average magnitude of errors between predicted and actual values in a regression model. It is calculated as the average of the absolute differences between the predicted and actual values. MAE is important because it provides a straightforward way to measure prediction accuracy in models.
4. Data storytelling is the process of translating complex data findings into a narrative that is easy to understand. Good data storytelling uses charts, graphs, and visualizations to highlight key insights and make the data more relatable and impactful. This technique helps decision-makers grasp complex data and make informed choices.
5. Hyperparameter tuning involves finding the best set of parameters that improve the performance of a machine learning model. Two common methods are:
   • Grid Search: Testing different combinations of hyperparameters to find the best performing model.
   • Random Search: Randomly selecting a subset of hyperparameter combinations for testing, which is faster than grid search.
6. Problem Scoping is the first phase of the AI Project Cycle. It involves identifying the problem to be solved, defining the objectives, understanding the stakeholders’ needs, and setting success criteria. This stage ensures that the AI solution aligns with business goals and stakeholder expectations.

---

8. Answer any 3 out of the given 5 questions.                                 (3*4)

1. K-Fold Cross Validation is a technique used to assess the performance of a machine learning model by splitting the dataset into K equal subsets (or “folds”). The model is trained on K-1 folds and tested on the remaining fold. This process is repeated K times, with each fold used exactly once as the test set. Common values for K are 5 or 10. It ensures that the model generalizes well to unseen data by reducing overfitting.
2. The steps in building a machine learning model are:
   • Data Collection: Acquiring relevant data from sources such as databases or APIs.
   • Data Preparation: Cleaning the data, handling missing values, and normalizing features.
   • Feature Engineering: Selecting or creating features that improve model performance.
   • Model Building: Choosing an algorithm (e.g., linear regression, decision trees) and training the model.
   • Model Evaluation: Assessing the model’s performance using metrics like accuracy, MAE, or RMSE.
   • Model Deployment: Deploying the model into production environments to generate predictions.
3. Classification and Regression are two types of machine learning algorithms:
   • Classification involves predicting discrete labels (e.g., spam or not spam). Common algorithms include logistic regression, decision trees, and SVMs.
   • Regression predicts continuous values (e.g., house prices). Algorithms include linear regression and polynomial regression.
4. RMSE Calculation for the data:
   Predicted: [10, 20, 30, 40]
   Actual: [12, 18, 29, 39]
   RMSE = √[(2² + 2² + 1² + 1²) / 4]
   RMSE = √[(4 + 4 + 1 + 1) / 4] = √10 / 4 = √2.5 ≈ 1.58
5. Exploratory Data Analysis (EDA) plays a crucial role in AI model building by helping data scientists understand the underlying patterns, structures, and relationships within the data before applying machine learning algorithms. Here are key ways in which EDA contributes to AI model building:

Identifying Data Patterns: EDA helps reveal important trends and patterns in the data, such as correlations between variables, distributions, and outliers. Understanding these patterns can guide the selection of features and algorithms for the model.

Handling Missing Data: EDA identifies missing values and inconsistencies within the dataset. By visualizing or summarizing missing data, it helps determine whether to fill, impute, or remove these gaps, ensuring the dataset is clean and ready for modeling.

Detecting Outliers and Anomalies: EDA helps detect outliers—data points that significantly differ from other observations. Outliers can affect model performance, so identifying and addressing them (either by removal or transformation) is essential for creating a robust model.

Feature Selection and Engineering: Through visualizations and statistical analysis, EDA allows for better feature selection by highlighting the most relevant variables. It also provides insights for feature engineering, enabling the creation of new features that improve model accuracy.