Skip to main content

Supervised vs unsupervised learning

Short answer

Supervised learning uses labeled data to predict outputs (e.g., spam detection), while unsupervised learning finds patterns in unlabeled data (e.g., customer segmentation). Supervised tasks include classification and regression; unsupervised includes clustering and association.

Differentiate supervised and unsupervised learning with suitable examples.

Supervised learning learns from training data that contains the correct output (class label) for each input, and builds a model to predict labels for new data. Unsupervised learning works on data without labels and finds hidden structure or groups by itself.

BasisSupervised learningUnsupervised learning
Training dataLabelled (input and known output)Unlabelled (input only)
GoalPredict a class or valueDiscover patterns and groups
SupervisionA "teacher" gives the correct answerNone
Main tasksClassification, regressionClustering, association, dimensionality reduction
EvaluationAccuracy on test set, easy to measureHarder, uses cluster quality measures
Number of classesKnown in advanceNot known (or chosen, e.g. k)
AlgorithmsDecision tree, Naive Bayes, KNN, SVM, neural networkk-means, hierarchical, DBSCAN, Apriori, PCA
ExamplePredict if an email is spam from emails already marked spam/not spamGroup customers into segments by buying behaviour

A third type, semi-supervised learning, uses a few labelled and many unlabelled examples.

✓ Answer checked by Claude Opus 5.5 (high effort)

All Data Mining chapter 3 questions →