clustering

name	clustering
description	Discover patterns in unlabeled data using clustering, dimensionality reduction, and anomaly detection
version	1.4.0
sasmp_version	1.4.0
bonded_agent	03-unsupervised-learning
bond_type	PRIMARY_BOND
parameters	{"required":[{"name":"X","type":"array","validation":"2D array, scaled numeric"}],"optional":[{"name":"n_clusters","type":"integer","default":null,"validation":"2 <= n <= 50"},{"name":"method","type":"string","default":"kmeans","validation":"[kmeans|dbscan|hierarchical|hdbscan]"}]}
retry_logic	{"strategy":"exponential_backoff","max_attempts":3,"base_delay_ms":1000}
logging	{"level":"info","metrics":["n_clusters","silhouette_score","execution_time"]}

Clustering Skill

Discover hidden patterns and groupings in unlabeled data.

Quick Start

from sklearn.cluster import KMeans
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import silhouette_score

# Always scale before clustering
scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)

# Cluster
kmeans = KMeans(n_clusters=5, random_state=42, n_init=10)
labels = kmeans.fit_predict(X_scaled)

# Evaluate
score = silhouette_score(X_scaled, labels)
print(f"Silhouette Score: {score:.4f}")

Key Topics

1. Clustering Algorithms

Algorithm	Best For	Key Params
K-Means	Spherical clusters	n_clusters
DBSCAN	Arbitrary shapes, noise	eps, min_samples
Hierarchical	Nested clusters	linkage
HDBSCAN	Variable density	min_cluster_size

from sklearn.cluster import KMeans, DBSCAN, AgglomerativeClustering
import hdbscan

algorithms = {
    'kmeans': KMeans(n_clusters=5, random_state=42),
    'dbscan': DBSCAN(eps=0.5, min_samples=5),
    'hierarchical': AgglomerativeClustering(n_clusters=5),
    'hdbscan': hdbscan.HDBSCAN(min_cluster_size=15)
}