MLD106 · Unsupervised Learning: Clustering and Segmentation

Categories: AI & Automation
Wishlist Share

About Course

Find natural groups in customers, products and transactions with clustering and dimensionality reduction, and turn them into segments a business can act on.

Advance · 8 hours · 4 lessons, a lab, an assessed assignment and a portfolio piece

Who it is for

Analysts, marketers and data practitioners who want data-driven customer or product segmentation.

Prerequisites

MLD103 and MLD105, or equivalent experience with pandas, visualization and scikit-learn pipelines.

What you will be able to do

  • Prepare features for clustering with appropriate scaling and transformations
  • Apply k-means, hierarchical clustering and DBSCAN and choose the number of clusters with evidence
  • Use PCA to reduce dimensions and visualize cluster structure
  • Profile, name and validate segments and recommend specific actions for each

Lessons

  • Lesson 1 · Clustering fundamentals and k-means
  • Lesson 2 · Other methods and dimensionality reduction
  • Lesson 3 · From clusters to business segments
  • Lesson 4 · Lab and assignment

Lab: RFM customer segmentation for an online store

Using two years of transactions from a small online gift shop, you will compute recency, frequency and monetary value per customer, cluster customers with k-means and compare the result with hierarchical clustering. You will profile each segment and propose a marketing action for it.

Free tools: Google Colab (free) with pandas, scikit-learn, matplotlib and seaborn

Assessed assignment

Segment customers, products or locations for a small business using the provided pet-store dataset or your own anonymized data. Deliver a notebook that justifies feature choices and scaling, compares at least two clustering methods, chooses the number of segments with quantitative and business evidence, and checks stability by re-running on a resample. Write a two-page segment playbook with a profile, a name, a size and a concrete, measurable action for each segment.

Criterion Weight What excellent work looks like
Feature preparation 20% Features fit the business goal, are transformed and scaled appropriately and the choices are explained.
Method and selection 30% At least two methods are compared, and the number of clusters is chosen using metrics, stability and interpretability.
Profiling and validation 25% Segments are profiled with clear statistics and visuals, and their stability is checked.
Actionable playbook 25% Each segment has a meaningful name, size and specific, measurable action a business could test.

Portfolio piece

A customer segmentation notebook and a segment playbook with profiles and recommended actions.

Self-paced, 12-month access. Questions are answered in writing through the student panel. A completion record is issued when you finish the course.

Next step

Continue with MLD107 · Model Evaluation, Tuning and Explainability · 8 hours · $56.00

Continue to the next course

Prefer the whole program? Program in Machine Learning & Applied Data Science · $378.00 (10% less than buying the courses separately)

Show More

What Will You Learn?

  • Prepare features for clustering with appropriate scaling and transformations
  • Apply k-means, hierarchical clustering and DBSCAN and choose the number of clusters with evidence
  • Use PCA to reduce dimensions and visualize cluster structure
  • Profile, name and validate segments and recommend specific actions for each

Course Content

MLD106 · Unsupervised Learning: Clustering and Segmentation (8 h)
Find natural groups in customers, products and transactions with clustering and dimensionality reduction, and turn them into segments a business can act on.

  • Lesson 1 · Clustering fundamentals and k-means
  • Lesson 2 · Other methods and dimensionality reduction
  • Lesson 3 · From clusters to business segments
  • Lesson 4 · Lab and assignment
  • MLD106 Assessment · Unsupervised Learning: Clustering and Segmentation
Scroll to Top