4. DBSCAN

Problem 4: Implement the following (30 Points):

Using the given code to generate a dataset:

from sklearn.datasets import make_blobs centers = [[1, 1], [-1, -1], [1, -1]] X, labels_true = make_blobs(n_samples=750, centers=centers, cluster_std=0.4, random_state=0) 

Use StandardScaler to normalize the dataset.

Perform DBSCAN on the dataset with parameters eps=0.3, min_samples=10

Print the estimated number of noise points (i.e., outliers) and Silhouette Coefficient for the clustering result.

Plot the data points in a scatter-plot style, coloring them to indicate their different membership (i.e., cluster association), like below (the figure is intentionally blurred to not reveal the correct answer):

Estimated number of clusters: 3