WorksheetsMachine Learning Quiz
Total questions: 86
Worksheet time: 43mins
Machine Learning mainly focuses on:
Writing manual rules
Learning patterns from data
Only storing data
Only cleaning data
Which of the following is NOT a type of Machine Learning?
Supervised
Unsupervised
Reinforcement
Typographic learning
In supervised learning, the model learns using:
Only unlabeled data
Labeled data
No data
Random data
Unsupervised learning works with:
Labeled data
Unlabeled data
Data with missing values
Only image data
Which of the following is an example of ML application?
Fans
Automatic email spam detection
Ceiling lights
None
Data preprocessing means:
Buying new data
Cleaning and preparing data
Ignoring data
Deleting all missing values
EDA stands for:
Easy Data Analysis
Exploratory Data Analysis
Extra Data Adjustment
Enhanced Data Algorithm
Which is NOT an ML application?
Weather prediction
Face recognition
Text translation
Manual typewriter
Semi-supervised learning uses:
Only labeled data
Both labeled and unlabeled data
No data
Only video data
Reinforcement Learning is based on:
Rewards and penalties
Labeled data
Image processing
Text analysis
Which comes first in ML workflow?
Model training
Data preprocessing
Model deployment
Model evaluation
A dataset is:
A single number
A collection of data
A machine
A graph
Missing values in data should be:
Ignored
Handled properly
Always deleted
Increased
Which term refers to understanding data using graphs?
Deployment
Prediction
Exploratory analysis
Modeling
Which is a supervised learning task?
K-Means
PCA
Regression
t-SNE
ML is a part of:
Chemistry
Artificial Intelligence
Geography
History
The target variable exists in:
Unsupervised learning
Supervised learning
Always
Never
ML models learn through:
Patterns
Fixed rules
Random selection
Guess
Label means:
A type of image
Correct output
A file name
Dataset size
ML helps in:
Automating decisions
Cooking food
Painting walls
None
Linear Regression predicts:
Categories
Continuous values
Words
Images
Logistic Regression is used for:
Regression only
Classification
Dimensionality reduction
Clustering
Decision Trees are used for:
Only regression
Only classification
Both classification and regression
Only clustering
Random Forest is a type of:
Single model
Ensemble model
Deep learning model
Reinforcement model
KNN depends on:
Distance
Trees
Layers
Rewards
SVM stands for:
Support Vector Machine
System Vector Mode
Simple Variable Machine
Standard Value Measure
Which algorithm uses Bayes’ theorem?
Random Forest
Naïve Bayes
K-Means
PCA
Gradient Descent is used for:
Minimizing error
Increasing error
Sorting data
Clustering
Boosting improves model performance by:
Reducing data
Combining weak learners
Increasing noise
Ignoring errors
AdaBoost stands for:
Adaptive Boosting
Added Boosting
Advanced Boosting
Auto Boosting
XGBoost is widely used because it is:
Slow
Very fast and accurate
Only for images
Very old
Accuracy measures:
Correct predictions
Model speed
Training time
Dataset size
Precision measures:
Correct positive predictions
Total data quality
Feature size
Negative classes
Recall measures:
All predicted positives
Correctly identified actual positives
False negatives
Dataset noise
F1 score is:
Average of precision & recall
Sum of precision & recall
Accuracy
Loss
ROC curve is used to evaluate:
Clustering
Classification models
Regression
PCA
Cross-validation helps to:
Reduce overfitting
Increase dataset
Change model type
Delete labels
Feature means:
Row
Column
Label
File
Training data is used to:
Evaluate model
Deploy model
Teach the model
Ignore errors
Test data is used to:
Train the model
Check model performance
Remove data
Normalize data
K-Means is a:
Supervised algorithm
Unsupervised algorithm
NLP algorithm
Deep learning model
Clustering means:
Grouping similar data
Predicting numbers
Translating text
Measuring accuracy
DBSCAN identifies clusters based on:
Distance & density
Labels
Features
Weights
DBSCAN identifies clusters based on:
Distance & density
Labels
Features
Weights
Hierarchical clustering produces:
A tree-like structure
A single number
A deep network
A regression line
Dimensionality reduction reduces:
Rows
Columns/features
Labels
Accuracy
PCA stands for:
Principal Component Analysis
Primary Component Arrangement
Partial Clustering Algorithm
Principal Category Analysis
PCA is mainly used for:
Clustering
Reducing features
Regression
Classification
LDA stands for:
Linear Data Algorithm
Linear Discriminant Analysis
Learning Data Arrangement
Label Distribution Analysis
t-SNE is used for:
Visualization
Regression
Boosting
Data cleaning
Unsupervised learning does NOT require:
Labels
Data
Features
Algorithm
The number of clusters in K-Means is:
Decided automatically
Given by the user
Always 2
Always 5
Outliers are handled well by:
K-Means
DBSCAN
SVM
Linear Regression
Similar data means:
Identical labels
Similar patterns
Same images
Same size
High-dimensional data means:
Many features
Many labels
Many rows
Noisy data
Semi-supervised learning requires:
100% labeled data
Few labeled + many unlabeled data
No data
Only images
Reinforcement Learning belongs to:
Unsupervised learning
Trial-and-error learning
Data mining
NLP
In clustering, groups are called:
Trees
Clusters
Labels
Predictions
PCA components are:
New transformed features
Labels
Predictions
Classes
Hierarchical clustering uses:
Dendrogram
Histogram
ROC curve
Precision curve
Unsupervised learning helps in:
Grouping
Labeling
Predicting exact value
Testing
Deep learning models are inspired by:
Human brain
Trees
Cars
Gravity
CNNs are best for:
Images
Audio
Text
Tables
RNNs are best for:
Independent data
Sequence data
Random data
Geographical data
NLP deals with:
Robots
Numbers
Human language
Weather
Text preprocessing includes:
Cleaning text
Writing text
Printing text
Encrypting text
Word embeddings represent words as:
Images
Vectors
Tables
Files
Transformer models work using:
Attention mechanism
Convolution
Reward system
Decision trees
BERT is used for:
Translation
Text understanding
Image processing
Sound detection
GPT stands for:
Generative Pre-trained Transformer
General Processing Technique
Graphical Pattern Tool
General Prediction Transformer
Deep learning requires:
More data
Less data
No data
Only labels
CNN stands for:
Convolutional Neural Network
Convolution Neural Node
Component Neural Net
Classified Neural Network
RNN stands for:
Recurrent Neural Network
Recursive Neural Node
Regional Neural Net
Random Neural Network
Transfer Learning helps by:
Starting from scratch
Using pre-trained models
Ignoring learning
Training endlessly
NLP is used in:
Chatbots
Washing machines
Refrigerators
Lights
Tokens in NLP are:
Sentences
Words or subwords
Tables
Features
Stop words are:
Important words
Common words to remove
Numbers
Special features
Word2Vec is used for:
Clustering
Creating word vectors
Cleaning text
Filtering labels
Sentiment analysis checks:
Grammar
Emotions in text
Model speed
Image size
Padding in DL is used to:
Match sequence lengths
Remove noise
Increase accuracy
Reduce time
NLP models convert text into:
Sound
Numbers
Images
Labels
Time series data is collected:
Randomly
Over time intervals
Without order
Only once
A popular time-series plot is:
Bar chart
Line chart
Pie chart
Scatter
Forecasting means:
Checking past
Predicting future
Deleting data
Visualizing
LSTM stands for:
Long Short-Term Memory
Large Short-Term Machine
Long Supervised Time Model
Linear System Time Model
LSTM is used for:
Sequence prediction
Static data
Clustering
Visualization
