Quiz on Unsupervised Learning

Introduction to Unsupervised Learning Answers 1. What does the Davies-Bouldin index specifically measure in clustering? A. The absolute number of clusters formed in the dataset B. The ratio of cluster separations to intra-cluster distances C. The average size of each cluster in the analysis D. The similarity between clusters based on their behavior Answer: The ratio of cluster separations to intra-cluster distances (B) The Davies-Bouldin index evaluates clustering quality by measuring the ratio of cluster separations to intra-cluster distances. 2. Which application of unsupervised learning focuses on grouping customers based on their behaviors? A. Anomaly Detection B. Data Visualization C. Customer Segmentation D. Image Compression Answer: Customer Segmentation (C) Customer Segmentation involves grouping customers based on their purchasing behavior and demographics. 3. Which of the following considers ground truth in clustering evaluation? A. Adjusted Rand index B. Feature Scaling C. Davies-Bouldin index D. Market Basket Analysis Answer: Adjusted Rand index (A) The Adjusted Rand index compares the clustering results with a ground truth. 4. What is a critical consideration when choosing an algorithm for unsupervised learning? A. The dataset's characteristics and goals of the analysis B. The number of features unrelated to the dataset C. The historical performance of different algorithms D. The color of the data points Answer: The dataset's characteristics and goals of the analysis (A) Choosing the right algorithm depends on the dataset's characteristics and the goals of the analysis. 5. Which unsupervised learning application aims to identify unusual events such as fraudulent transactions? A. Recommendation Systems B. Image Compression C. Anomaly Detection D. Data Visualization Answer: Anomaly Detection (C) Anomaly Detection focuses on identifying unusual events, such as fraudulent transactions. 6. What is the primary goal of unsupervised learning? A. To train algorithms using input-output pairs B. To classify data into pre-defined categories C. To discover hidden patterns in unlabeled data D. To enhance the accuracy of predictions Answer: To discover hidden patterns in unlabeled data (C) Unsupervised learning aims to analyze data without prior labels to find patterns. 7. Which of the following is NOT a common application of unsupervised learning? A. Dimensionality reduction B. Customer segmentation C. Anomaly detection D. Sentiment analysis Answer: Sentiment analysis (D) Sentiment analysis typically requires labeled data, whereas the other applications do not. 8. How does K-means clustering determine the number of clusters? A. It is determined automatically by the algorithm B. It uses a fixed number of clusters by default C. It determines clusters based on data density D. It requires the user to specify the number of clusters Answer: It requires the user to specify the number of clusters (D) K-means clustering necessitates a pre-defined number of clusters, stated as K. 9. What does Principle Component Analysis (PCA) primarily do? A. Analyzes market basket relationships B. Detects outliers in a dataset C. Clusters data into distinct groups D. Transforms data to reduce dimensions while preserving variance Answer: Transforms data to reduce dimensions while preserving variance (D) PCA reduces the number of variables while keeping the most significant information about variance. 10. Which clustering algorithm is known for its ability to identify clusters of arbitrary shapes? A. K-means clustering B. Hierarchical clustering C. DBSCAN D. Gaussian Mixture Models Answer: DBSCAN (C) DBSCAN allows for the detection of clusters that do not form spherical shapes, unlike K-means. 11. What does the silhouette score measure in the context of clustering? A. The time efficiency of the clustering algorithm B. The overall population of the dataset C. The quality of clustering by comparing cluster similarity D. The variance within individual clusters Answer: The quality of clustering by comparing cluster similarity (C) The silhouette score evaluates how closely a data point resembles its own cluster compared to others. 12. Which of the following methods is used for association rule learning? A. K-means clustering B. t-SNE visualization C. Hierarchical clustering D. Market basket analysis Answer: Market basket analysis (D) Market basket analysis is a primary example of association rule learning, revealing relationship patterns. 13. What advantage does t-SNE provide when working with highdimensional data? A. It preserves local distances between data points B. It reduces data size through linear transformations C. It employs a fixed number of dimensions regardless of data D. It creates probabilistic models for clustering Answer: It preserves local distances between data points (A) t-SNE is particularly effective for visualizing clusters by maintaining the local structure of the data.