WorksheetsAI and Machine Learning Worksheet Questions (Transcribed)
Total questions: 50
Worksheet time: 25mins
Artificial Intelligence mainly focuses on which human intelligence processes?
Cooking and cleaning
Learning, reasoning, and self-correction
Only data storage
Hardware design
Which type of AI is currently used in voice assistants like Siri?
General AI
Super AI
Narrow AI
Strong AI
General AI is best described as:
AI for gaming only
AI that can perform one task well
AI with human-like intelligence across tasks
AI used only in research labs
Why does General AI not exist yet?
Lack of internet
High computational and philosophical challenges
No programming languages
No data available
Machine Learning is a subset of:
Data Science
Deep Learning
Artificial Intelligence
Statistics
Machine Learning systems improve mainly by:
Manual updates
Experience from data
Hardware changes
Internet speed
Which of the following is NOT a type of Machine Learning mentioned?
Supervised Learning
Unsupervised Learning
Reinforcement Learning
Transfer Learning
In supervised learning, data used is:
Random
Unlabeled
Semi-labeled
Labeled
Which algorithm is commonly used in supervised learning?
K-Means
Apriori
Linear Regression
HMM
Spam email detection is an example of:
Regression
Clustering
Classification
Dimensionality reduction
Predicting house prices comes under:
Classification
Regression
Clustering
Association
In unsupervised learning, the system:
Uses labeled output
Learns with teacher
Finds patterns on its own
Always predicts numbers
Customer segmentation is an example of:
Supervised learning
Unsupervised learning
Reinforcement learning
Semi-supervised learning
Which algorithm is used for association rule learning?
SVM
K-Means
Apriori
Logistic Regression
Dimensionality reduction mainly helps to:
Increase features
Remove labels
Reduce features and simplify data
Semi-supervised learning uses:
Only labeled data
Only unlabeled data
Both labeled and unlabeled data
Reinforcement signals only
Reinforcement learning works based on:
Predefined rules
Trial and error
Labeled datasets
Clustering
The main goal of reinforcement learning is to:
Reduce loss
Maximize total reward
Minimize data
Increase features
Which is a key element of reinforcement learning?
Token
Agent
Corpus
Vocabulary
Q-learning mainly uses:
Neural weights
Probability table
Q-table
Decision tree
SARSA differs from Q-learning because it:
Uses deep learning only
Uses next action from current policy
Needs full environment knowledge
Is rule-based
Deep Learning is inspired by:
Statistics
Human brain
Internet
Hardware
Deep Learning mainly uses:
Decision trees
Single-layer models
Artificial Neural Networks with many layers
Rule engines
Backpropagation is used to:
Collect data
Improve model performance
Generate rules
Remove noise
Which layer gives final prediction in neural networks?
Input layer
Hidden layer
Output layer
Attention layer
Generative AI is mainly used to:
Predict labels only
Store data
Create new content
Compress files
ChatGPT is an example of:
Rule-based model
Statistical model
Transformer-based model
Clustering model
Generative AI learns patterns from:
Small datasets
Hardware instructions
Large datasets
Only rules
Which model uses predefined human-written rules?
Statistical model
Neural model
Rule-based model
Generative model
Naive Bayes belongs to which model type?
Rule-based
Statistical
GAN
Transformer
GAN consists of:
Encoder and decoder
Generator and discriminator
Agent and environment
Token and corpus
VAEs are mainly used to:
Classify data
Compress and recreate data
Label data
Remove noise
Transformer models are efficient in handling:
Images only
Small data
Large text data
Numerical data only
Diffusion models generate images by:
Classification
Rule matching
Step-by-step noise removal
Decision trees
Discriminative models mainly learn:
Data distribution
Decision boundary
Noise patterns
Random rules
Which is a discriminative model?
GAN
VAE
Logistic Regression
Diffusion model
Discriminative models are best suited for:
Data generation
Classification and prediction
Image creation
NLP mainly helps computers to:
Store data
Understand human language
Design hardware
Improve networks
NLP acts as a bridge between:
AI and ML
Human language and machine language
Hardware and software
Data and models
Tokenization means:
Removing stopwords
Converting text into tokens
Translating language
Summarizing text
Stopword removal helps to:
Add new words
Remove unimportant words
Create vocabulary
Increase accuracy always
Bag of Words is used in:
Data labeling
Feature extraction
Output generation
Model evaluation
NLU mainly focuses on:
Generating text
Understanding meaning
Translating text
Speech synthesis
NLG is responsible for:
Understanding intent
Generating human-like text
Token creation
Feature extraction
Named Entity Recognition is used to:
Count words
Identify entities like names and places
Translate text
Remove symbols
A token is best defined as:
Complete dataset
Smallest unit of text
Set of unique words
Collection of documents
Vocabulary refers to:
All documents
All sentences
Set of unique tokens
Set of models
Corpus means:
Single sentence
Token list
Large collection of text data
Feature vector
Large vocabulary size generally leads to:
Less memory usage
Faster processing
More memory usage
No effect
Correct relationship is:
Tokens → Corpus → Vocabulary
Vocabulary → Tokens → Corpus
Corpus → Tokens → Vocabulary
Tokens → Vocabulary → Corpus
