subtopic

The unsupervised workflow and preparing data for distances

Write the 7-step workflow; then run one demo showing how unscaled features dominate distances. Concepts: Workflow; feature choice; scaling (Standard / MinMax / Robust); log transforms; mixed data (glimpse) Project A: Demo: k-means on raw vs log+scaled RFM features — completely different segments. Project B: Demo: day profiles in raw MW (groups by level/season) vs each day divided by its mean (groups by shape).

Builds on: E11 · Build: EDA for Segmentation (Project A without the target)