Intermediate and Advanced Questions on Machine

Learning, NLP, and Robotics

Multiple Choice Questions

1. What is the primary goal of machine learning as described in Chapter 5?
a) To explicitly program computers for specific tasks
b) To enable computers to learn from data and improve over time
c) To simulate human emotions in computers
d) To replace all human decision-making processes
Answer: b) To enable computers to learn from data and improve over time
2. In the context of machine learning, what is the purpose of the testing phase?
a) To train the algorithm with labeled data
b) To evaluate the model’s performance on unseen data
c) To preprocess the data for training
d) To define the target attributes
Answer: b) To evaluate the model’s performance on unseen data
3. Which of the following is a key feature of machine learning according to Chapter
5?
a) It relies solely on manual coding
b) It is a data-driven technology
c) It avoids the use of historical data
d) It is unrelated to data mining
Answer: b) It is a data-driven technology
4. How is supervised learning classified in machine learning?
a) Based on the absence of labeled data
b) Based on feedback in the form of rewards or penalties
c) Based on the presence of labeled data

1
 d) Based on clustering similar data points
Answer: c) Based on the presence of labeled data
5. Which task is NOT typically associated with supervised learning?
a) Binary classification
b) Regression modeling
c) Clustering
d) Ensembling
Answer: c) Clustering
6. In regression analysis, what does the dependent variable represent?
a) The variable that influences the analysis
b) The variable that is being forecasted
c) The error term in the model
d) The constant term in the equation
Answer: b) The variable that is being forecasted
7. What does the slope coefficient (β1 ) represent in a linear regression model?
a) The value of y when x is 0
b) The change in y for a one-unit change in x
c) The error term in the model
d) The number of features in the model
Answer: b) The change in y for a one-unit change in x
8. Which algorithm is primarily used for classification problems in supervised learning?
a) Linear Regression
b) Logistic Regression
c) K-Means Clustering
d) Hierarchical Clustering
Answer: b) Logistic Regression
9. In a Support Vector Machine (SVM), what is the role of the hyperplane?
a) To preprocess the input data
b) To segregate data into classes
c) To reduce the dimensionality of data
d) To cluster similar data points
Answer: b) To segregate data into classes

2
10. What distinguishes a probabilistic classification model from other classification
models?
a) It predicts continuous values
b) It predicts a probability for each class
c) It requires unlabeled data
d) It uses rewards and penalties
Answer: b) It predicts a probability for each class
11. What is the primary goal of unsupervised learning?
a) To predict categorical outputs
b) To find patterns in unlabeled data
c) To optimize rewards in an environment
d) To train with labeled data
Answer: b) To find patterns in unlabeled data
12. Which of the following is an example of a clustering algorithm?
a) Logistic Regression
b) K-Means Clustering
c) Support Vector Machine
d) Linear Regression
Answer: b) K-Means Clustering
13. What is the purpose of association mining in unsupervised learning?
a) To predict continuous values
b) To identify items that frequently occur together
c) To classify data into categories
d) To optimize decision-making
Answer: b) To identify items that frequently occur together
14. In hierarchical clustering, what structure is created to represent the data?
a) A hyperplane
b) A dendrogram
c) A decision boundary
d) A regression line
Answer: b) A dendrogram
15. What is a key characteristic of reinforcement learning?

3
 a) It uses labeled data for training
b) It relies on rewards and penalties
c) It clusters data into groups
d) It predicts continuous values
Answer: b) It relies on rewards and penalties
16. In deep learning, what is the role of the hidden layers in a neural network?
a) To provide the final output
b) To learn complex features from data
c) To preprocess the input data
d) To store the training data
Answer: b) To learn complex features from data
17. How does a Convolutional Neural Network (CNN) differ from a standard neural
network?
a) It uses loops to memorize past inputs
b) It is designed for processing 2D data like images
c) It requires labeled data for clustering
d) It predicts categorical outputs only
Answer: b) It is designed for processing 2D data like images
18. What is a key advantage of CNNs in image processing?
a) They require manual feature extraction
b) They automatically extract features from images
c) They are limited to grayscale images
d) They cannot handle complex shapes
Answer: b) They automatically extract features from images
19. Why are Recurrent Neural Networks (RNNs) suitable for tasks like next word pre-
diction?
a) They process data in a single pass
b) They have a memory of previous inputs
c) They are designed for image processing
d) They require unlabeled data
Answer: b) They have a memory of previous inputs
20. What is the role of Long Short-Term Memory (LSTM) in RNNs?
a) To reduce the number of layers

4
 b) To enhance memory for longer sequences
c) To preprocess input data
d) To classify categorical data
Answer: b) To enhance memory for longer sequences
21. In Natural Language Processing (NLP), what is a corpus?
a) A single sentence in a text
b) A collection of documents or text files
c) A tokenized word
d) A labeled dataset
Answer: b) A collection of documents or text files
22. What is the first step in the NLP pipeline?
a) Word Tokenization
b) Sentence Segmentation
c) Stemming
d) Lemmatization
Answer: b) Sentence Segmentation
23. What is the main difference between stemming and lemmatization in NLP?
a) Stemming produces a root word with meaning
b) Lemmatization produces a root word with meaning
c) Stemming is used for part-of-speech tagging
d) Lemmatization removes stop words
Answer: b) Lemmatization produces a root word with meaning
24. What is the purpose of removing stop words in NLP?
a) To increase the complexity of the text
b) To remove non-essential words to speed up processing
c) To tag parts of speech
d) To translate text into another language
Answer: b) To remove non-essential words to speed up processing
25. In NLP, what does Named Entity Recognition (NER) detect?
a) The sentiment of a text
b) Specific entities like person names or locations
c) The grammatical structure of sentences

5
 d) The root words of tokens
Answer: b) Specific entities like person names or locations
26. Which NLP application predicts whether a review is positive, negative, or neutral?
a) Language Translation
b) Sentiment Analysis
c) Automatic Summarization
d) Spam Filtering
Answer: b) Sentiment Analysis
27. What is the role of a bilingual machine translation system in NLP?
a) To translate between multiple languages
b) To translate between two specific languages
c) To summarize text documents
d) To detect named entities
Answer: b) To translate between two specific languages
28. What is a key challenge in building a question-answering system in NLP?
a) Processing numerical data
b) Handling lexical gaps and ambiguity
c) Clustering similar data points
d) Predicting continuous values
Answer: b) Handling lexical gaps and ambiguity
29. What is the primary goal of computer vision as described in Chapter 6?
a) To process textual data
b) To enable computers to interpret visual data
c) To translate languages
d) To cluster data points
Answer: b) To enable computers to interpret visual data
30. How does image processing differ from computer vision?
a) Image processing involves cognitive operations like object recognition
b) Image processing focuses on enhancing or extracting information from images
c) Computer vision focuses on image enhancement
d) Image processing requires labeled data

6
 Answer: b) Image processing focuses on enhancing or extracting information from
images
31. In robotics, what distinguishes a robot from a machine?
a) A robot uses an open-loop control system
b) A robot uses a closed-loop control system
c) A machine is autonomous
d) A robot requires human intervention
Answer: b) A robot uses a closed-loop control system
32. According to Asimov’s Three Laws of Robotics, what is the first law?
a) Robots must protect themselves
b) Robots must follow human instructions
c) Robots must never harm humans
d) Robots must optimize their tasks
Answer: c) Robots must never harm humans
33. What is the primary function of a robot sensor?
a) To execute tasks autonomously
b) To measure the robot’s condition and environment
c) To translate human commands
d) To cluster data points
Answer: b) To measure the robot’s condition and environment
34. Which type of sensor is NOT mentioned in Chapter 7?
a) Light Sensor
b) Sound Sensor
c) Temperature Sensor
d) Motion Sensor
Answer: d) Motion Sensor
35. In human-robot interaction (HRI), what is remote interaction?
a) Interaction where humans and robots are co-located
b) Interaction where humans and robots are spatially separated
c) Interaction requiring labeled data
d) Interaction using only visual data
Answer: b) Interaction where humans and robots are spatially separated

7
36. What is the goal of path planning in robotics?
a) To process sensory data
b) To find the shortest obstacle-free path
c) To translate human commands
d) To cluster similar objects
Answer: b) To find the shortest obstacle-free path
37. What are the three key components of autonomous robotic systems?
a) Sensing, clustering, and actuation
b) Perception, decision, and actuation
c) Classification, regression, and planning
d) Tokenization, lemmatization, and translation
Answer: b) Perception, decision, and actuation
38. In robotics, what role do actuators play?
a) To process sensory data
b) To convert energy into movement
c) To translate human commands
d) To cluster data points
Answer: b) To convert energy into movement
39. Which of the following is an example of a service robot mentioned in Chapter 7?
a) ASIMO
b) Roomba
c) Pepper
d) All of the above
Answer: b) Roomba
40. What is a key limitation of human vision compared to computer vision?
a) Limited to the visible spectrum
b) Ability to process 2D data
c) High memory capacity
d) Resistance to optical illusions
Answer: a) Limited to the visible spectrum
41. In NLP, what is the purpose of POS tagging?
a) To translate text

8
 b) To assign grammatical categories to words
c) To remove stop words
d) To segment sentences
Answer: b) To assign grammatical categories to words
42. Which of the following is a disadvantage of stemming in NLP?
a) It produces meaningful root words
b) It may produce root words without meaning
c) It requires labeled data
d) It is used for translation
Answer: b) It may produce root words without meaning
43. In machine learning, what is ensembling?
a) Combining multiple models to improve predictions
b) Reducing the number of features
c) Clustering similar data points
d) Translating text data
Answer: a) Combining multiple models to improve predictions
44. What is the role of the intercept (β0 ) in a linear regression model?
a) The change in y for a one-unit change in x
b) The value of y when x is 0
c) The error term in the model
d) The number of classes in the model
Answer: b) The value of y when x is 0
45. In SVM, what are support vectors?
a) The central points of clusters
b) The extreme points that define the hyperplane
c) The hidden layers in a neural network
d) The stop words in a text
Answer: b) The extreme points that define the hyperplane
46. What is a key application of clustering mentioned in Chapter 5?
a) Predicting continuous values
b) Recommending products based on past searches
c) Translating languages

9
 d) Classifying categorical data
Answer: b) Recommending products based on past searches
47. In robotics, what is the purpose of navigation sensors?
a) To process textual data
b) To help robots determine their position and path
c) To translate human commands
d) To cluster similar objects
Answer: b) To help robots determine their position and path
48. What is the primary difference between linear and non-linear SVM?
a) Linear SVM uses labeled data, non-linear does not
b) Linear SVM is for linearly separable data
c) Non-linear SVM is used for regression
d) Linear SVM requires more layers
Answer: b) Linear SVM is for linearly separable data
49. In deep learning, what does the term ”deep” refer to?
a) The number of input features
b) The number of hidden layers
c) The size of the training data
d) The complexity of the output
Answer: b) The number of hidden layers
50. Which of the following is a use case for NLP mentioned in Chapter 6?
a) Image classification
b) Speech recognition
c) Path planning
d) Clustering
Answer: b) Speech recognition
51. In robotics, what is the purpose of proximate interaction in HRI?
a) Interaction where humans and robots are spatially separated
b) Interaction where humans and robots are co-located
c) Interaction requiring rewards and penalties
d) Interaction using only visual data
Answer: b) Interaction where humans and robots are co-located

10
Fill in the Blank Questions
1. Machine learning is a subset of ________ that involves training algorithms to
learn from data.
2. In supervised learning, the training data must contain the correct answer, known
as a ________.
3. The ________ phase in machine learning evaluates the model’s performance on
unseen data.
4. In regression analysis, the ________ variable is the one being forecasted.
5. The ________ term in a linear regression model represents random variation
not explained by the model.
6. Logistic regression predicts ________ values that lie between 0 and 1.
7. In unsupervised learning, ________ is used to group objects with similar pat-
terns.
8. ________ clustering divides data into non-hierarchical groups based on a cen-
troid.
9. Reinforcement learning uses ________ and penalties to improve the agent’s
performance.
10. In deep learning, the ________ layer is responsible for receiving input data.
11. A ________ Neural Network is designed for processing 2D data like images.
12. In NLP, a ________ is a collection of documents or text files.
13. The process of breaking a sentence into individual words is called ________.
14. ________ is used to normalize words into their base form with meaning in NLP.
15. In NLP, ________ words like ”is” and ”the” are removed to speed up processing.
16. Named Entity Recognition (NER) detects entities like ________ names or lo-
cations.
17. ________ analysis in NLP predicts whether a review is positive, negative, or
neutral.
18. In robotics, a ________ control system aligns output without human interven-
tion.
19. The first of Asimov’s Three Laws of Robotics states that robots must never ________
humans.
20. In robotics, ________ planning involves finding the shortest obstacle-free path.

True/False Questions
1. Machine learning algorithms require explicit programming for each task.
2. Supervised learning uses unlabeled data to train models.

11
 3. Logistic regression can be used for both classification and regression tasks.
4. Clustering is a supervised learning technique.
5. Recurrent Neural Networks (RNNs) have a memory of previous inputs.
6. In NLP, stemming always produces a root word with meaning.
7. A robot is distinguished from a machine by its closed-loop control system.
8. Human-robot interaction (HRI) only involves proximate interactions.
9. Computer vision aims to mimic human vision but can overcome its limitations.
10. Path planning in robotics requires a map of the environment.

Answers to Fill in the Blank and True/False Questions

Fill in the Blank Answers
1. Artificial Intelligence
2. Target
3. Testing
4. Dependent
5. Error
6. Probabilistic
7. Clustering
8. Partitioning
9. Rewards
10. Input
11. Convolutional
12. Corpus
13. Tokenization
14. Lemmatization
15. Stop
16. Person
17. Sentiment
18. Closed-Loop
19. Harm
20. Path

12
True/False Answers
1. False
2. False
3. False
4. False
5. True
6. False
7. True
8. False
9. True
10. True

13