Clustering is defined as "an unsupervised process that partitions data into groups (clusters) based on similarity without preassigned labels." K-Means is the canonical unsupervised clustering algorithm, iteratively assigning points to K centroids to minimize within-cluster variance. This makes K-Means the simplest and most direct choice for grouping data in an anomaly-detection context.
=========

Phase II: Data Understanding is dedicated to identifying data requirements, collecting initial data, assessing data quality, and verifying that necessary datasets are accessible and fit for modeling. Determining what data you need and confirming access are the core activities of this phase .

A spam filter must decide between exactly two categories-spam or not spam-making it a binary (or binomial) classification task. The CPMAI Glossary defines binary classification as "a classification task where data is categorized into one of two classes (e.g., spam vs. not spam)."
=========

Phase III: Data Preparation focuses on constructing and documenting reusable data pipelines-including training and inference pipelines-alongside cleansing, augmentation, and labeling tasks to prepare data for modeling . This is where teams build the end-to-end data preparation workflows for AI solutions such as chatbots.

Under Domain I: Evaluating AI Applications and Patterns, CPMAI instructs practitioners to "Integrate multiple AI patterns for comprehensive applications" when solutions span more than one cognitive pattern.
Treating a multi-pattern system as a single, cohesive project ensures aligned data streams, shared infrastructure, and unified governance.
=========