Advanced Concepts of Modeling in AI
MCQ
Artificial Intelligence (417) – MCQs
1. Which of the following is the umbrella term that covers both Machine Learning and Deep Learning?
- (a) Machine Learning
- (b) Deep Learning
- (c) Artificial Intelligence
- (d) Data Science
Answer: (c) Artificial Intelligence
2. What is the primary characteristic of a Supervised Learning model?
- (a) It works on unlabeled data.
- (b) It learns by trial and error based on a reward mechanism.
- (c) It requires labeled data for training.
- (d) It is based on a predefined set of rules.
Answer: (c) It requires labeled data for training.
3. Which type of AI model is based on rules and instructions defined by a developer?
- (a) Learning-Based Approach
- (b) Reinforcement Learning
- (c) Unsupervised Learning
- (d) Rule-Based Approach
Answer: (d) Rule-Based Approach
4. A model that predicts a continuous value like temperature or price is a:
- (a) Classification Model
- (b) Regression Model
- (c) Clustering Model
- (d) Association Model
Answer: (b) Regression Model
5. What is the main advantage of Neural Networks mentioned in the document?
- (a) They are based on a simple decision tree approach.
- (b) They require minimal data for training.
- (c) They are able to automatically extract data features.
- (d) They only work on numerical data.
Answer: (c) They are able to automatically extract data features.
6. The process of attaching meaning or tags to data is known as:
- (a) Data Processing
- (b) Data Mining
- (c) Data Labeling
- (d) Data Extraction
Answer: (c) Data Labeling
7. Which of the following is a subset of Machine Learning?
- (a) Artificial Intelligence
- (b) Supervised Learning
- (c) Rule-Based Approach
- (d) All of the above
Answer: (b) Supervised Learning
8. In an Artificial Neural Network, which layer is responsible for processing the data using weights and biases?
- (a) Input Layer
- (b) Hidden Layer
- (c) Output Layer
- (d) Data Layer
Answer: (b) Hidden Layer
9. A supermarket uses an AI model to group its customers based on their purchase history to send targeted offers. What kind of model is this?
- (a) Supervised Learning
- (b) Reinforcement Learning
- (c) Classification
- (d) Clustering
Answer: (d) Clustering
10. Which learning approach is characterized by machines learning from feedback in a trial-and-error method?
- (a) Supervised Learning
- (b) Unsupervised Learning
- (c) Reinforcement Learning
- (d) Classification
Answer: (c) Reinforcement Learning
Artificial Intelligence (417) – MCQs (Questions 11–20)
11. The core processing in an Artificial Neural Network occurs in which layer?
- (a) Input Layer
- (b) Hidden Layer
- (c) Output Layer
- (d) All layers
Answer: (b) Hidden Layer
12. A model is trained to classify handwritten digits (0–9). This is an example of:
- (a) Unsupervised Learning
- (b) Regression
- (c) Classification
- (d) Clustering
Answer: (c) Classification
13. Which of the following is a drawback of the Rule-Based Approach?
- (a) It requires vast amounts of data.
- (b) It is too complex to implement.
- (c) The learning is static and does not adapt to changes.
- (d) It cannot be used for chatbots.
Answer: (c) The learning is static and does not adapt to changes.
14. In the context of data, what are "features"?
- (a) Rows of a table
- (b) The output of a model
- (c) Columns of a table
- (d) The final prediction
Answer: (c) Columns of a table
15. What is the difference between a training dataset and a testing dataset?
- (a) Training data is unlabeled, while testing data is labeled.
- (b) Training data is used to teach the model, and testing data is used to evaluate its accuracy.
- (c) Training data contains only features, and testing data contains only labels.
- (d) There is no difference; they are interchangeable.
Answer: (b) Training data is used to teach the model, and testing data is used to evaluate its accuracy.
16. A model that predicts whether an email is "spam" or "not spam" is a type of:
- (a) Regression Model
- (b) Clustering Model
- (c) Classification Model
- (d) Association Model
Answer: (c) Classification Model
17. In the example of predicting a coin's currency based on its weight, what is the "feature"?
- (a) The weight
- (b) The currency
- (c) The coin's name
- (d) The number of coins
Answer: (a) The weight
18. An AI model that discovers patterns in an unlabeled dataset of dog images, such as clustering them by color or size, uses which approach?
- (a) Supervised Learning
- (b) Reinforcement Learning
- (c) Unsupervised Learning
- (d) Rule-Based Approach
Answer: (c) Unsupervised Learning
19. Which of the following is a characteristic of a Learning-Based Approach?
- (a) It is based on static rules.
- (b) It learns from explicit programming.
- (c) It adapts to changes in data.
- (d) It does not require any data.
Answer: (c) It adapts to changes in data.
20. What is a "perceptron"?
- (a) A type of neural network layer
- (b) A machine learning algorithm
- (c) A simplified model of how an AI makes a decision
- (d) A type of data
Answer: (c) A simplified model of how an AI makes a decision
Artificial Intelligence (417) – MCQs (Questions 21–30)
21. Which type of model is used to predict a car's approximate selling price based on parameters like fuel type and years of service?
- (a) Classification
- (b) Regression
- (c) Clustering
- (d) Association
Answer: (b) Regression
22. What is the role of the Output Layer in an Artificial Neural Network?
- (a) To process data and pass it to the hidden layers.
- (b) To perform calculations with weights and biases.
- (c) To present the final processed data to the user.
- (d) To acquire data and feed it to the network.
Answer: (c) To present the final processed data to the user.
23. In the context of a supermarket, which unsupervised learning method would be used to find a relationship between customers buying bread and also buying butter?
- (a) Classification
- (b) Clustering
- (c) Regression
- (d) Association
Answer: (d) Association
24. Which type of learning is analogous to a child learning to swim on his own without a teacher?
- (a) Supervised Learning
- (b) Unsupervised Learning
- (c) Reinforcement Learning
- (d) Rule-Based Learning
Answer: (b) Unsupervised Learning
25. What do you need to train a supervised learning model?
- (a) Unlabeled data
- (b) Only features
- (c) Only labels
- (d) Labeled data
Answer: (d) Labeled data
26. What is the main drawback of a Rule-Based AI model?
- (a) It is too expensive to implement.
- (b) It fails to learn from its mistakes.
- (c) It requires continuous human intervention.
- (d) It can only handle a single type of data.
Answer: (b) It fails to learn from its mistakes.
27. When an AI model trains itself to perform tasks with vast amounts of data, it falls under which category?
- (a) Machine Learning
- (b) Deep Learning
- (c) Rule-Based Approach
- (d) Supervised Learning
Answer: (b) Deep Learning
28. Which layer in an Artificial Neural Network does not perform any processing?
- (a) Input Layer
- (b) Hidden Layer
- (c) Output Layer
- (d) Both (a) and (c)
Answer: (d) Both (a) and (c)
29. Object classification in Deep Learning uses powerful algorithms to identify and label objects within:
- (a) An image
- (b) A spreadsheet
- (c) A text document
- (d) A database
Answer: (a) An image
30. The process of a neural network finding the right output by adjusting weights based on the error is known as:
- (a) Trial and Error
- (b) Backpropagation
- (c) Forward Propagation
- (d) Activation
Answer: (b) Backpropagation
Artificial Intelligence (417) – MCQs (Questions 31–40)
31. What is the key difference between Clustering and Classification?
- (a) Classification uses unlabeled data, while clustering uses labeled data.
- (b) Classification assigns objects to predefined classes, while clustering finds similarities and groups objects.
- (c) Classification is a supervised model, while clustering is a reinforcement model.
- (d) Clustering requires more data than classification.
Answer: (b) Classification assigns objects to predefined classes, while clustering finds similarities and groups objects.
32. What is the goal of Reinforcement Learning?
- (a) To identify relationships in unlabeled data.
- (b) To predict a continuous value.
- (c) To make a series of decisions that maximize a reward.
- (d) To classify data into discrete categories.
Answer: (c) To make a series of decisions that maximize a reward.
33. Anomaly detection, such as flagging a sudden spike in a heart rate, is an example of which type of model?
- (a) Classification
- (b) Regression
- (c) Machine Learning
- (d) Reinforcement Learning
Answer: (c) Machine Learning
34. In the example of a spam email filter, what serves as the "labels" during the training phase?
- (a) The words in the email
- (b) The sender's information
- (c) The classification of emails as either "spam" or "legitimate"
- (d) The email attachments
Answer: (c) The classification of emails as either "spam" or "legitimate"
35. What is the purpose of a testing dataset?
- (a) To train the model with new data.
- (b) To check for errors in the code.
- (c) To evaluate the accuracy of the trained model.
- (d) To find hidden patterns.
Answer: (c) To evaluate the accuracy of the trained model.
36. Which of the following is an example of a Classification problem?
- (a) Predicting the price of a house.
- (b) Predicting the number of days a patient will stay in a hospital.
- (c) Predicting whether a patient will have a short or long hospital stay.
- (d) Predicting a city's average temperature for the next month.
Answer: (c) Predicting whether a patient will have a short or long hospital stay.
37. What type of data is used for training a Regression model?
- (a) Categorical data
- (b) Discrete data
- (c) Continuous data
- (d) Unlabeled data
Answer: (c) Continuous data
38. Which type of model is used by OTT platforms like Netflix to recommend movies based on a user's watch history?
- (a) Supervised Learning
- (b) Regression
- (c) Clustering
- (d) Reinforcement Learning
Answer: (c) Clustering
39. The structure of an Artificial Neural Network is inspired by:
- (a) The human brain and nervous system
- (b) The human digestive system
- (c) A computer's hardware components
- (d) A mathematical formula
Answer: (a) The human brain and nervous system
40. In a supervised learning example, what does a model learn from the training data?
- (a) To create new features
- (b) To identify new patterns without guidance
- (c) To apply the knowledge to test data
- (d) To define the rules itself
Answer: (c) To apply the knowledge to test data
Artificial Intelligence (417) – MCQs (Questions 41–50)
41. Which of the following is a primary type of AI model mentioned in the document?
- (a) Predictive-based
- (b) Reward-based
- (c) Learning-based
- (d) Both (b) and (c)
Answer: (c) Learning-based
42. What is a "label" in the context of data?
- (a) A column of a table
- (b) The name of the dataset
- (c) A tag that gives meaning to data
- (d) The algorithm used in the model
Answer: (c) A tag that gives meaning to data
43. A Convolutional Neural Network (CNN) is a type of:
- (a) Machine Learning Model
- (b) Rule-Based Model
- (c) Deep Learning Algorithm
- (d) Reinforcement Learning Model
Answer: (c) Deep Learning Algorithm
44. When a model predicts a discrete value, such as "hot" or "cold" weather, it is using a:
- (a) Regression Model
- (b) Classification Model
- (c) Association Model
- (d) Clustering Model
Answer: (b) Classification Model
45. The "funnel type approach" described for AI, ML, and DL suggests that:
- (a) DL has more applications than AI.
- (b) All ML applications are also DL applications.
- (c) All AI applications are also ML applications.
- (d) DL has a very specific set of applications, a subset of ML, which is a subset of AI.
Answer: (d) DL has a very specific set of applications, a subset of ML, which is a subset of AI.
46. Which learning approach would be used to analyze bank data for suspicious transactions without a predefined definition of what is "suspicious"?
- (a) Supervised Learning
- (b) Unsupervised Learning
- (c) Regression
- (d) Classification
Answer: (b) Unsupervised Learning
47. What is a key function of the Hidden Layer in a neural network?
- (a) To acquire input data.
- (b) To give the final output.
- (c) To perform computations.
- (d) To act as a user interface.
Answer: (c) To perform computations.
48. What is the main difference between a Rule-Based Approach and a Learning-Based Approach?
- (a) Rule-based is static, while learning-based is adaptive.
- (b) Rule-based is more efficient for large datasets.
- (c) Learning-based is based on predefined rules.
- (d) There is no significant difference.
Answer: (a) Rule-based is static, while learning-based is adaptive.
49. What are "weights" and "biases" used for in a neural network?
- (a) To define the final output.
- (b) To acquire input data.
- (c) To perform computations within the hidden layers.
- (d) To create a user interface.
Answer: (c) To perform computations within the hidden layers.
50. What does Deep Learning enable software to do?
- (a) To learn from a minimal amount of data.
- (b) To train itself to perform tasks with vast amounts of data.
- (c) To follow a predefined set of rules.
- (d) To predict discrete values only.
Answer: (b) To train itself to perform tasks with vast amounts of data.
Assertion and Reasoning
Artificial Intelligence (417) – Assertion & Reason (Questions 1–10)
Directions: Choose the correct option.
- (a) Both Assertion (A) and Reason (R) are true, and Reason (R) is the correct explanation of Assertion (A).
- (b) Both Assertion (A) and Reason (R) are true, but Reason (R) is NOT the correct explanation of Assertion (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 models require labeled data for training.
Reason (R): Labeled data acts as a guide, providing the model with a clear relationship between features and outcomes.
Answer: (a) Both A and R are true, and R is the correct explanation of A.
2. Assertion (A): A model predicting a house price based on its size is a Regression model.
Reason (R): Regression models are used to predict discrete values.
Answer: (c) Assertion (A) is true, but Reason (R) is false.
3. Assertion (A): Anomaly detection is a key application of Unsupervised Learning.
Reason (R): Anomaly detection models look for outliers and unusual patterns in unlabeled data.
Answer: (a) Both A and R are true, and R is the correct explanation of A.
4. Assertion (A): A Rule-based AI model is adaptive and can handle new exceptions it was not explicitly programmed for.
Reason (R): Rule-based models follow a static set of instructions predefined by a developer.
Answer: (d) Assertion (A) is false, but Reason (R) is true.
5. Assertion (A): Deep Learning is a subset of Machine Learning.
Reason (R): Deep Learning algorithms are based on artificial neural networks and require vast amounts of data for self-training.
Answer: (a) Both A and R are true, and R is the correct explanation of A.
6. Assertion (A): Reinforcement Learning models learn by trial and error to maximize a reward.
Reason (R): This learning approach is beneficial for complex problems where pre-existing data is insufficient.
Answer: (b) Both A and R are true, but R is NOT the correct explanation of A.
7. Assertion (A): A model classifying an email as "spam" or "not spam" is a Classification model.
Reason (R): Classification models predict a continuous value.
Answer: (c) Assertion (A) is true, but Reason (R) is false.
8. Assertion (A): The hidden layers of a neural network are where the core processing of data takes place.
Reason (R): The input layer and output layer of an ANN perform all the computational tasks.
Answer: (c) Assertion (A) is true, but Reason (R) is false.
9. Assertion (A): Unsupervised Learning is analogous to a teacher-student relationship.
Reason (R): Unsupervised Learning models work on unlabeled data without any guidance.
Answer: (d) Assertion (A) is false, but Reason (R) is true.
10. Assertion (A): The primary purpose of the testing dataset is to train the model.
Reason (R): The training dataset is used to evaluate the accuracy of the model.
Answer: (d) Assertion (A) is false, but Reason (R) is false.
Artificial Intelligence (417) – Assertion & Reason (Questions 11–20)
Directions: Choose the correct option.
- (a) Both Assertion (A) and Reason (R) are true, and Reason (R) is the correct explanation of Assertion (A).
- (b) Both Assertion (A) and Reason (R) are true, but Reason (R) is NOT the correct explanation of Assertion (A).
- (c) Assertion (A) is true, but Reason (R) is false.
- (d) Assertion (A) is false, but Reason (R) is true.
11. Assertion (A): The "weights" in a neural network are adjusted during the learning process.
Reason (R): The adjustment of weights helps to reduce the error and find the right output.
Answer: (a) Both A and R are true, and R is the correct explanation of A.
12. Assertion (A): Reinforcement Learning is a type of Supervised Learning.
Reason (R): It involves a reward system that provides feedback, which is a form of supervision.
Answer: Both Assertion (A) and Reason (R) are false. Reinforcement Learning is a separate learning approach that uses rewards, not labeled supervision. (If your school uses only options a–d, this question is incorrectly framed.)
13. Assertion (A): Clustering is a type of Unsupervised Learning.
Reason (R): It is used to group data points into clusters based on their similarities without using predefined classes.
Answer: (a) Both A and R are true, and R is the correct explanation of A.
14. Assertion (A): A model predicting whether a patient will have a short or long hospital stay is a Regression model.
Reason (R): The output of this prediction is a continuous value.
Answer: (d) Assertion (A) is false, but Reason (R) is true.
15. Assertion (A): The Input Layer of an Artificial Neural Network (ANN) does not process data.
Reason (R): Its sole function is to acquire data and feed it to the next layer.
Answer: (a) Both A and R are true, and R is the correct explanation of A.
16. Assertion (A): An Association model finds relationships between variables in a dataset.
Reason (R): An example is recommending bread to a customer who buys butter.
Answer: (a) Both A and R are true, and R is the correct explanation of A.
17. Assertion (A): Deep Learning requires only a small amount of data to train a model effectively.
Reason (R): Deep Learning enables software to train itself with vast amounts of data.
Answer: (d) Assertion (A) is false, but Reason (R) is true.
18. Assertion (A): A perceptron is a simplified model of how an AI makes a decision.
Reason (R): It works by summing weighted inputs and a bias, then comparing the result to a threshold.
Answer: (a) Both A and R are true, and R is the correct explanation of A.
19. Assertion (A): A spam email filter is a good example of a Learning-Based model.
Reason (R): It adapts and improves its accuracy over time as it encounters new types of spam.
Answer: (a) Both A and R are true, and R is the correct explanation of A.
20. Assertion (A): The process of attaching meaning or tags to data is called Data Labeling.
Reason (R): Data Labeling is a crucial step in preparing data for Unsupervised Learning.
Answer: (c) Assertion (A) is true, but Reason (R) is false.
