Unsupervised Learning & Clustering

Course outline and learning expectations

This is a tutor-led course. The outline shows what your tutor will cover; teaching materials and examinations are provided directly to enrolled students.

What you will learn

• Explain unsupervised learning and recognise suitable problems without labelled outcomes
• Prepare and scale tabular data for clustering
• Build and interpret K-Means models and select an appropriate number of clusters
• Apply hierarchical clustering and interpret a dendrogram
• Use PCA to reduce dimensions and visualise structure in data
• Use basic anomaly detection to flag unusual observations for further review
• Develop customer or mobile money user segments from sample transaction data
• Communicate cluster profiles, limitations and practical recommendations

Modules

1. 1

   Foundations of Unsupervised Learning

   Understand how algorithms find structure in data without labelled targets

   Unsupervised versus supervised learningCommon problem typesClustering workflowResponsible use of unlabelled data
2. 2

   K-Means Clustering Foundations

   Build customer groups using distance-based clustering

   Feature selectionData cleaningFeature scalingK-Means algorithmCentroids and cluster assignments
3. 3

   K-Means Evaluation and Interpretation

   Choose and explain useful clusters for business decisions

   Elbow methodSilhouette scoreCluster profilingVisualising clustersLimits of K-Means
4. 4

   Hierarchical Clustering

   Create nested groups and interpret relationships between observations

   Agglomerative clusteringDistance metricsLinkage methodsDendrogramsComparing hierarchical clustering with K-Means
5. 5

   Dimensionality Reduction with PCA

   Explain and apply PCA to simplify and visualise multivariable data

   Variance and componentsStandardisationExplained varianceTwo-dimensional visualisationInterpreting PCA carefully
6. 6

   Anomaly Detection Basics

   Flag unusual records for investigation without treating them as proof of fraud

   Outliers and anomaliesIsolation Forest basicsThresholdsFalse positivesFraud-screening workflowEthical and privacy considerations
7. 7

   Market Segmentation Case Study

   Turn cluster results into clear segment profiles and practical recommendations

   Segmentation objectivesCustomer featuresPersona developmentSegment validationCommunicating insights
8. 8

   Capstone: Customer and Mobile Money Segmentation

   Segment sample transaction data into behaviour groups and present evidence-based findings

   Problem definitionData preparationClustering and evaluationPCA visualisationSegment profilesRecommendationsFinal presentation