stud
bud
Sign up
Log in
Home
Library
Explore
SAT
More
AI Quiz Flashcards | StudBud
Explore
/
AI Quiz
AI Quiz
79 terms
·
StudBud Administration
·
15 plays
Imported from external platform
Flashcards
79 cards
F
Practice quiz
P
Host live
L
Questions
79
Answers
1
Which of the following would be an example of Artificial General Intelligence (AGI) if it existed?
program that can accurately identify and tag frinds in photos on platforms
Algorithm optimize logistics and supply routes within predefined parameters
AI that diagnoses diseases, invents theories, writes novels, and much more
A voice-activated assistant that play music set timers and weather updates
2
What is the basis of item-based collaborative filtering?
Finding similar items based on user behavior
Clustering users based on demographics
Using a bag of words model
based on item popularity
3
What is depth first search?
search algorithm that explores search space by going deep before backtrack
search algorithm that always chooses the shortest path first
search algorithm that explores the search space in a breadth-first manner
search algorithm that uses heuristics to guide the search
4
What is the main goal of the NLP pipeline
To convert images into text data
To transofmr textual data into numerical vectors for ML and DL analysis
To transform textual data into audio signals
To create a new language model from scratch
5
Which of the follwoing is an example of convolution used for image processing
file format conversion
Color correction
smoothing using an average box filter
image compression
6
What does the R² score measure in Regression?
How well the model fits the spread of the datapoints
The correlation between variables
The accuracy of the model predictions
The residuals squared
7
What is the purpose of Linear Regression?
To find a straight line (model) to map the data as close as possible
To classify data into different categories
To analyze the relationship between variables
To predict future values based on historical data
8
Which of the following isn't a type of graph
Weighted Graph
Relational Database
Undirected Graph
Directed Graph
9
What is the purpose of the learning cycle in deep learning
Calculate the error or loss of the model during the forward pass.
Tweak the weights of the model based on the error or loss calculated
Adjust the activation functions of the model during forward pass
Calculate the gradients of the model during the forward pass
10
What kind of machine learning problem is the "iris dataset" typically used for?
Classification
Regression
Clustering
Anomaly detection
11
What is the primary purpose of the turing test?
to determine the accuracy of decision trees
to assess the speed of deep learning models
to evvaluate the efficiency of recursive functions
To measure a machine's ability to exhibit intelligent behavior
12
Content-based recommender systems make recommendations based on what?
Features of the content item
Random Selection
The popularity of items
User collaboration
13
Why might a popularity-based system by problematic
It can lead to a narrow range of recommendations
It only recommends new items
It assumes all users have the same taste
It requires too much data
14
What is Dijkstra's algorithm?
A search algorithm that uses heuristics to guide the search
An algorithm that finds the shortest path between nodes in a graph
search algorithm that explores search space by going deep before backtrack
15
Collaborative filtering in recommender systems is based on:
Item popularity
User collaboration
Random forest algorithm
Content similarity
16
Named Entity Recognition (NER) is used to:
Identify and categorize key information in text into predefined categories.
Create new names for fictional entities in creative writing
Predict the future trends in language usage
Translate named entities from one language to another
17
What does a sparse graph have
Short paths
lots of edges
few vertices
few edges
18
Mogelijk examenvraag:
vraag me niet het antwoord
ik heb geen idee wat antwoord moet zijn of de vraag
19
What contributed to the recent boom in deep learning technology
Increased funcing for AI research
Introduction of new programming languages
Advances in computational power and availability of large dataset
Adoption of traditional AI techniques
20
which of these is an example of a problem that could be effectively solved using a recursive a
Counting number of words in a large document by scanning beginning to end
algorithm to reverse the caracters in a string, such as turning Hoi in ioh
Implementing a search function to find a specific word in a paragraph
21
What does Narrow AI refer to?
AI's designed for specific tasks can't work beyond their programmed domain.
AI system that surpass human intelligenc in all fields including creativity
hypothetical AI stage that autonomously improves beyond human understanding
AI systems capable of performing any intellectual task that a human can
22
What is the purpose of date augmentation in training CNNs
To reduce the number of layers in the network
To enlarge the dataset and prevent overfitting
To convert all training images to black and white
To decrease the dataset size for faster training
23
Which of the following scenarios best describes Super AI
self-driving car that can navigate city streets better than a human driver
digital assistant that can manage emails better than any human secratary
An AI capable of beating the world champion in chess
24
How is a user profile typically built in content-based recommender system?
By gathering data on user preferences
By clustering items only
By using a decision tree
By random sampling
25
What is the goal of a classification algorithm in machine learning
to devide data points into different categories
to cluster similar data points together
to find the mean of the data
to predict a continuous value
26
What is the concept of "bagging" in the context of Random Forest?
Selecting the best feature for splitting
Combining different algorithms
Choosing the best decision tree
Randomy selecting samples to create multiple datasets
27
Why might the BoW model be problematic
It cannot analyze text longer than a sentence.
It changes the meaning of the original text
It can only analyze words in English
It results in high dimensional sparse data vectors
28
What is the purpose of ensemble learning in machine learning?
To combine multiple models to improve predictions
To simplify the model
To increase the speed of computations
To reduce the number of features nedded
29
What is the risk of making a model too complex in Polynomial Regression
Making the model too complex
Having too few parameters
Creating a biased model
Underfitting the data
30
What is the first step in Regression Modeling
Importing & wrangling data
Calculating the performance of the model
Splitting the data into train and test set
Fitting the data to a model
31
What does cosine similarity measure?
The angle between two non-zero vectors
The correlation between variables
The distance between two points
The causality between events
32
What is the term for the boundary that a classification algorithm tries to determine
hyperlink
hyperplane
hyperparameter
hyperspace
33
In unsupervised learning, what is the significance of weighted rating?
It is used to rank items based on popularity
it ranks the features by importance
it is used to calculate the cost function
it determines the learning rate
34
What is a significant advantage of GAP over Flattening in CNNs
It allows for more complex patterns
It enhances the resolution of the output image
It simplifies the model and speeds up computation
It increases the accuracy of edge detection
35
What is the outpout of the following recursisve function if the input (n) is 5
20
16
10
15
36
What is the role of an optimizer in the learning cycle?
adjust the activation fucntions of the mdoel during the forward pass.
Calculate the gradients of hte model during the forward pass
help gradient descent algorithm to find a good set of weights and biases
Calculate the error or loss of the model during the forward pass
37
What is the goal of model performance measurement in Regression?
To maximize the explained variance
To minimize the mean absolute error
To minimize the RMS-error
To maximize the R2 score
38
What is informed search?
A search algorithm that always chooses the shortest path first
a search algorithm that explores the search space in a breadth-first manner
A search algorithm that uses heuristics to guide the search
a search algorithm that does not have any information about the goal state
39
What is the role of the learning rate in model training?
To set the initial values of the model weights
To control how much the model changes in response to the cost function
To decide the type of activation function to use in the network
To determine the number of layers in the network
40
In a decision tree, what is a leaf node?
The most important feature
The starting point of the tree
A node that does not split further
The maximum depth of the tree
41
What is the primary purpose of the Gradient Descent algorithm in machine learning
To increase the complexity of the model
To prevent overfitting
To classify data points
To find a set of weights and biases which minimize the cost
42
what does Tokenization in NLP genarally aim to achieve
Compressing text files to save space
Translating text from one language to another
Splitting text into smaller units like words or sentences
Encrypting text for secure communication
43
Consider the following recursive function to calculate the factorial of a number n: what does factorial(5) return
15
120
5
25
44
The process of adjusting which parameters to improve a model's performance is known as?
Feature selection
Model evaluation
Error reduction
Parameter tuning
45
What does convolution in computer vision primarily involve
Applying a sliding window technique
changing image formats
Scaling images to different sizes
Rotating images for better alignment
46
What problem does recursion offer an elegant solution to in coding?
Decision trees
Counting files in a folder
Path finding code
Complex problems with repeating patterns or structures
47
What is the purpose of dropout in deep learning?
Decrease the learning rate of the model during training
Adjust the activation functions of the model during training
Increase the number of neurons in the model during training
Prevent overfitting by dropping out percentage of neurons during training
48
What issue can arise with new users or items in collaborative filtering?
High variance
Overfitting
Cold Start
Underfitting
49
What is the concept behind a layered neural network?
Each layer performs a specific mathematical operation on the input data
each layer stores a different type of information about the input data
Each layer acts as a (more complex) feature extractor
Each layer represents a different class or category of the input data
50
A machine learning, what is the process of adjusting a model to better fit the data known as?
learning (fitting)
regulatization
overfitting
underfitting
51
What does a CNN (convolutional neural network) learn during training?
The optimal size for image resizing
Predefined kernels
The weights in the kernels
The best image formats for processing
52
What is the main goal when calculating user similarity in collaborative filtering
To find similar users based on their ratings and preferences
To cluster content into categories
To determine the most popular content
To identify the best content for all users
53
What is a node in a graph?
A series of actions
An object with a certain state
A planning
A goal-state
54
what are the components of csp solving
possible objects and their states
variables, domains and constraints
constraint graphs
rules to obey in planning
55
What is a significant advantage of Word2Vec over BoW
Word2Vec uses simple count metrics instead of neural networks.
Word2Vec is much slower and less efficient
Word2Vec can only analyze single words, not phrases
Word2Vec provides dense word embeddings with semantic meanings
56
In machine leraning
Supervised learning
semi-supervised learning
reinforcement learning
unsupervised learning
57
What is the purpose of a CV benchmark competition like Imagenet?
To apply convolutional operations on an image
To downsample a feature map
to drive innovation in computer vision
To create a labeled database
58
which of the following statements about recursion is correct?
recursive fucntion is fucntion that calls itself within its own definition
recursive function always executes faster than equivalent iterative functio
Recursion can only be used for mathematical computations
Recursion does not require the use of a termination condition
59
What is the 'Elbow Method' used for in K-means clustering?
Choosing the optimal number of clusters
Selecting the best features
Reducing dimensionality of the dataset
Optimizing the learning rate
60
What is one-hot encoding used for?
Reducing computational cost
Normalizing features
Labeling categories of movies
Encoding textual data
61
What is the main purpose of using the sobel operator in image processing?
color enhancement
Noise reduction
Edge detection
Image scaling
62
What distinguishes feature extraction form fine tuning in transfer learning
FT only output layer is trained with pre-trained model weights frozen
FE trains new models from scratch; fine-tuning involves modifying
Feature Extraction and Fine Tuning are identical
FE only output layer is trained with pre-trained model weights frozen
63
What is the main difference between stemming and lemmatization?
Stemming changes text into foreign language, while lemmatization keep same
Color enhancement
There is no difference; both are the same
Stemming chops off word while lemmatization considers morphological analysi
64
What is the function of biases in a neural network?
to set the activation threshold
to prevent overfitting
to decide the learning rate
To initialize the weights
65
What is the main difference between the average box filter and the gaussian filter in image processing
Box filter changes image dimensions, where Gaussian filter maintains them.
The box filter uses a fixed threshold, whereas the gaussianfilter adapts
There is no difference; both filters perform identical operations
box filter averages neighboring pixel, the gaussian uses Weighted averaging
66
What is the purpose of a 'Head' in transfer learning?
To extract features form an image
To apply convolutional operations on an image
To flatten a feature map
to add a custom classifier to the base model
67
Which of the following is an example of narrow AI
A robot that has learned to cook any dish by watching videos online
A chatbot designed to provide customer support for a specific product
AI system that can write novels that are indistinct from written by humans
A computer program that can autonomoysly develop new AI algorithms
68
What is property of a pre-trained network?
They limit the amount of needed labeled data
They need less computational power during run-time
They are light-weight
They simplify the run-time computation
69
What is a common step in text preprocessing for NLP?
Removing stopwords to reduce noise
Converting all text into uppercase
Increasing the font size of the text
Changing the text color
70
What is the aim of using a confusion matrix in classification tasks
To optimize the model parameters
To visualize the algorithm's performance
To estimate the complexity of the model
to compute the accuracy of the model
71
What are constraints in constraint satisfaction problems CSP
Constraint graphs
rules to obey in planning
possible objets and their states
variables domains constraints
72
How is recursion described as an elegant solution in problem-solving?
it uses loops to solve problems iteratively
It involves complex coding techniques
it simplifies complex problems into smaller instances
it requires less computational power
73
What is the effect of pooling in a CNN
Enhances the color of the image
Increases the number of layers
Reduces the size of the image
Converts the image into grayscale
74
What is the term for a dataset used to evaluate the fit of a model?
Validation set
Inference set
Training set
Test set
75
What search method uses a queue date structure?
Deapht-First search
Breadth-First search
76
What do you call a model that is created in a single pass over the data
Iterative learner
batch learner
online learner
one-shot learner
77
What is the search strategy used in linear Regression?
Dijkstra
Gradient descent
Decision Tree
Random Forest
78
What is the link between nodes in a graph called?
Edges
Loops
Vertices
Paths
79
Which of the following is a way to represent a graph
Adjacency List
Adjacency Matrix
Queue
Stack
