1. . Unsupervised learning is:
(A) Training models with labeled data only
(B) A type of machine learning where the model is trained on unlabeled data to find patterns or structure
(C) Compressing labeled datasets
(D) Backup only
2. . The main goal of unsupervised learning is to:
(A) Encrypt patterns
(B) Discover hidden patterns, clusters, or relationships in data
(C) Compress datasets
(D) Backup only
3. . Clustering in unsupervised learning is:
(A) Compressing clusters
(B) Encrypting clusters
(C) Grouping similar data points into clusters
(D) Backup only
4. . K-Means is:
(A) Backup only
(B) A supervised learning algorithm
(C) Encryption algorithm
(D) A clustering algorithm that partitions data into K clusters based on similarity
5. . Hierarchical clustering builds:
(A) A flat label set
(B) A hierarchy of clusters using either agglomerative or divisive methods
(C) An encryption tree
(D) Backup only
6. . Dimensionality reduction in unsupervised learning is:
(A) Compressing datasets
(B) Encrypting features
(C) Reducing the number of input variables to simplify data and improve analysis
(D) Backup only
7. . Principal Component Analysis (PCA) is used to:
(A) Reduce dimensionality while preserving variance in data
(B) Encrypt principal components
(C) Compress features only
(D) Backup only
8. . Independent Component Analysis (ICA) is used for:
(A) Separating a multivariate signal into independent non-Gaussian components
(B) Encrypting signals
(C) Compressing signals
(D) Backup only
9. . Anomaly detection in unsupervised learning is:
(A) Encrypting anomalies
(B) Identifying unusual data points that differ significantly from the majority
(C) Compressing anomalies
(D) Backup only
10. . DBSCAN is:
(A) A density-based clustering algorithm that identifies clusters and outliers
(B) A supervised regression algorithm
(C) Encryption method
(D) Backup only
11. . Autoencoders are:
(A) Encrypting autoencoders
(B) Neural networks used for unsupervised feature learning and dimensionality reduction
(C) Compressing neural networks
(D) Backup only
12. . Silhouette score in clustering measures:
(A) How well data points are clustered, indicating cohesion and separation
(B) Encrypting clusters
(C) Compressing clusters
(D) Backup only
13. . Unsupervised learning is useful for:
(A) Compressing patterns
(B) Encrypting marketing data
(C) Market segmentation, anomaly detection, pattern discovery, and dimensionality reduction
(D) Backup only
14. . Cosine similarity is used in unsupervised learning to:
(A) Encrypt similarity
(B) Measure similarity between two vectors
(C) Compress vectors
(D) Backup only
15. . Agglomerative clustering starts with:
(A) One big cluster
(B) Each data point as its own cluster and merges them iteratively
(C) Encrypting clusters
(D) Backup only
16. . Divisive clustering starts with:
(A) All data points in one cluster and splits them iteratively
(B) Each point as a cluster
(C) Encrypting divisions
(D) Backup only
17. . Feature scaling is important in unsupervised learning because:
(A) Compressing features
(B) Encrypting features
(C) Distance-based algorithms like K-Means are sensitive to feature magnitudes
(D) Backup only
18. . The elbow method is used to:
(A) Determine the optimal number of clusters in K-Means
(B) Encrypt K-values
(C) Compress clusters
(D) Backup only
19. . t-SNE is a technique used for:
(A) Visualizing high-dimensional data in 2D or 3D
(B) Encrypting high-dimensional data
(C) Compressing features
(D) Backup only
20. . The main purpose of unsupervised learning is to:
(A) Discover hidden structure, patterns, or relationships in unlabeled data
(B) Encrypt all data
(C) Compress datasets
(D) Backup only