Supervised vs unsupervised learning exam review

This flashcard set reviews the key concepts of supervised and unsupervised learning, including definitions, differences, examples, and applications in artificial intelligence and machine learning.

MasonDragon·32 flashcards·32 questions
collegecomputer_scienceai_ml
0
Known
1 / 32
0
Learning
Front

What is supervised learning?

Tap to flip
Back

A type of machine learning where the model is trained on labeled data. It learns from input-output pairs.

Tap to flip
Got it
Still learning

Quiz(32 questions)

Question 1 of 32

1. What is the primary goal of unsupervised learning?

Terms in this Study Set(32)

Fundamentals of Supervised Learning(16)

What is supervised learning?

A type of machine learning where the model is trained on labeled data. It learns from input-output pairs.

What are labeled data?

Data that includes both input features and the corresponding output labels. Examples: (input: images, output: categories).

True or False: Supervised learning can be used for regression tasks.

True. It can predict continuous outcomes, like prices or temperatures.

Name two common algorithms for supervised learning.

- Linear Regression - Decision Trees

Fill in the blank: Supervised learning requires ______ data.

labeled

Cause → Effect: What happens when you train a supervised model?

The model learns to map inputs to outputs, improving prediction accuracy.

What is classification in supervised learning?

A task where the model predicts categorical labels. Example: email spam detection.

How does overfitting occur?

When a model learns noise in the training data too well, failing on new data.

What is the goal of supervised learning?

To develop a predictive model that generalizes well to unseen data.

True or False: Supervised learning can only be applied to structured data.

False. It can also handle semi-structured and unstructured data with appropriate preprocessing.

Comparison: Supervised vs. Unsupervised Learning

Supervised uses labeled data; Unsupervised uses unlabeled data.

What is a training set?

A subset of data used to train the model, consisting of input-output pairs.

How is a validation set used?

To tune hyperparameters and prevent overfitting during training.

What is the output of a regression model?

A continuous value. Example: predicting house prices in dollars.

Give an example of a supervised learning application.

Predicting customer churn based on historical data.

True or False: Supervised learning uses only categorical outputs.

False: Supervised learning can use categorical outputs (classification) and continuous outputs (regression), depending on the problem type.

Fundamentals of Unsupervised Learning(16)

What is unsupervised learning?

A type of machine learning where the model learns patterns from unlabelled data.

True or False: Unsupervised learning uses labeled data.

False. It relies on unlabelled data to find patterns or groupings.

Example of unsupervised learning.

Clustering customers into segments based on purchasing behavior without predefined labels.

What is clustering?

A technique in unsupervised learning that groups similar data points together.

Fill in the blank: The most common algorithm for clustering is ______.

K-means.

Difference between clustering and association.

Clustering groups data by similarity; association identifies rules of relationships between variables.

What are dimensionality reduction techniques?

Methods to reduce the number of features in a dataset while retaining essential information. Examples include PCA and t-SNE.

What is PCA?

Principal Component Analysis; a technique for reducing dimensionality by transforming to a new set of variables.

Cause → Effect: High-dimensional data → ______.

Increased computational complexity and overfitting risk.

What does anomaly detection do?

Identifies rare items or events in a dataset that differ significantly from the majority.

Example of anomaly detection.

Fraudulent transactions in banking systems detected as outliers in spending patterns.

True or False: Unsupervised learning can be used for data visualization.

True. It helps visualize high-dimensional data in lower dimensions, aiding understanding.

What is hierarchical clustering?

A method that builds a tree of clusters by merging or splitting them based on similarity.

Applications of unsupervised learning.

Market segmentation, social network analysis, and organizing computing clusters.

What is feature extraction?

The process of transforming raw data into a set of usable features for analysis.

True or False: Unsupervised learning guarantees accurate predictions.

False. It finds patterns, but predictions are not the primary goal.

Questions in this Study Set(32)

1. What is the primary goal of unsupervised learning?

A.To find hidden patterns in unlabelled data
B.To predict outcomes based on labeled data
C.To classify data into predefined categories
D.To improve the accuracy of supervised models

2. What type of data does supervised learning utilize?

A.Labeled data
B.Unlabeled data
C.Partially labeled data
D.Random data

3. Which of the following is an example of clustering?

A.Grouping similar books in a library
B.Identifying the price of a stock based on historical data
C.Predicting the weather based on temperature trends
D.Classifying emails as spam or not spam

4. Which of the following algorithms is commonly used for classification tasks in supervised learning?

A.K-Means Clustering
B.Linear Regression
C.Support Vector Machines
D.Principal Component Analysis

5. True or False: K-means is a clustering algorithm.

A.True
B.False
C.Sometimes
D.Depends on the context

6. True or False: Supervised learning can address both classification and regression problems.

A.True
B.False
C.Only classification
D.Only regression

7. Which statement is NOT true about hierarchical clustering?

A.It creates a tree-like structure of clusters
B.It can be used for large datasets efficiently
C.It merges clusters based on similarity
D.It can split clusters to refine groupings

8. How does a model benefit from being trained on a training set?

A.It becomes aware of all possible inputs
B.It learns to generalize from input-output pairs
C.It can only predict seen data
D.It memorizes exact data points

9. What is the main use of dimensionality reduction techniques?

A.To increase the number of features in a dataset
B.To simplify models by reducing features while preserving data integrity
C.To cluster data into segments
D.To assign labels to unlabelled data

10. What is the main goal of supervised learning?

A.To cluster data
B.To develop a predictive model
C.To visualize complex data
D.To analyze text data

11. What does PCA stand for?

A.Principal Component Analysis
B.Primary Cluster Algorithm
C.Pattern Classification Application
D.Predictive Clustering Analysis

12. Which of the following is NOT a type of supervised learning task?

A.Classification
B.Regression
C.Clustering
D.Prediction

13. Which of the following is an application of unsupervised learning?

A.Market segmentation
B.Credit scoring
C.Image classification
D.Sentiment analysis

14. What happens during the validation phase of supervised learning?

A.The model is trained
B.The model is tested on unseen data
C.Hyperparameters are tuned
D.The model learns new patterns

15. What does anomaly detection focus on?

A.Identifying patterns in regular data
B.Finding rare items or events that differ significantly from the majority
C.Classifying data into multiple groups
D.Analyzing trends over time

16. How can overfitting negatively impact a supervised learning model?

A.It improves accuracy on training data
B.It ensures better predictions on all data types
C.It reduces prediction accuracy on new data
D.It simplifies the model

17. True or False: Unsupervised learning is only used in clustering tasks.

A.True
B.False
C.Sometimes
D.Only in academic settings

18. Fill in the blank: In supervised learning, the model learns from a series of ______.

A.Features
B.Noisy data
C.Input-output pairs
D.Unlabeled samples

19. Which algorithm is commonly used for clustering?

A.K-means
B.Linear regression
C.Decision trees
D.Random forest

20. What indicates that a supervised learning model is performing well?

A.High accuracy on training data only
B.Consistent predictions on unseen data
C.Complex model structure
D.Low data usage during training

21. What is the relationship between high-dimensional data and overfitting?

A.High-dimensional data reduces overfitting risk
B.High-dimensional data can increase overfitting risk
C.High-dimensional data has no effect on overfitting
D.Overfitting only occurs in supervised learning

22. Which of these is a common metric used to evaluate a supervised learning model’s performance?

A.Silhouette score
B.Mean Absolute Error
C.Dunn index
D.Covariance

23. Which of the following is NOT a technique for dimensionality reduction?

A.PCA
B.t-SNE
C.Clustering
D.Autoencoders

24. True or False: Supervised learning can only be applied to numerical data.

A.True
B.False
C.Only continuous data
D.Only categorical data

25. What is the purpose of feature extraction?

A.To generate raw data
B.To convert data into a format suitable for analysis
C.To increase the number of features
D.To eliminate all features

26. What is the purpose of a test set in supervised learning?

A.To train the model
B.To measure the model's accuracy
C.To tune parameters
D.To preprocess data

27. True or False: Unsupervised learning can improve interpretability of data.

A.True
B.False
C.Only in small datasets
D.Only with labeled data

28. Which scenario best exemplifies a supervised learning application?

A.Grouping customers by spending habits
B.Predicting housing prices based on features
C.Discovering patterns in text data
D.Clustering similar movies together

29. What is the main difference between clustering and association?

A.Clustering finds relationships; association groups similar items
B.Clustering groups similar items; association finds rules between variables
C.Both are the same
D.Clustering is a supervised method; association is unsupervised

30. In regression tasks, what type of output is expected?

A.Categorical
B.Binary
C.Continuous
D.Nominal

31. Which of the following best describes the process of feature extraction?

A.Transforming raw data into usable features for analysis
B.Grouping similar data points together
C.Identifying rules of relationships between variables
D.Eliminating all outliers from a dataset

32. Which of the following best describes the role of labeled data in supervised learning?

A.It provides input-output pairs for model training.
B.It is used only for validation purposes.
C.It helps in unsupervised learning tasks.
D.It is irrelevant to machine learning.

Related Study Sets

Create Your Own Study Set

Upload a PDF, paste your notes, or describe a topic – AI generates flashcards, quizzes and more in seconds.