AI Foundations Through Data

Beginner / Python 3 / 10 lessons

Grouping similar records differs from predicting a label for a new record. This course follows how numerical representations reveal patterns and support decisions.

• Who it is for: You know arithmetic and averages and can read short Python examples.
• Topics: Features, distances, hierarchical clustering, K-means, fuzzy clustering, Bayes classification, nearest neighbors, evaluation, and missing values.
• Examples: Small teaching datasets that expose the calculations, not reported performance of real services.
• Goal: Separate training from evaluation and examine why a result appears successful.

Every lesson is free to read. Before memorizing a formula, identify the data being compared and the assumptions being made.

Lessons

1. 01 What Are We Predicting or Grouping?Distinguish tasks by their goals and whether data has labels.
2. 02 Build Features and DistancesSee how numerical representation changes the meaning of similarity.
3. 03 Merge Nearby GroupsSee how linkage defines distances between clusters.
4. 04 Move Centers with K-MeansAlternate assignment to a center with recomputing cluster means.
5. 05 Choose Useful ClustersCompare initial centers and group counts, then evaluate the results.
6. 06 Handle Points Near a BoundaryDistinguish fuzzy cluster membership from calibrated probability.
7. 07 Update Beliefs with BayesCalculate a prior, likelihood, and posterior in a small example.
8. 08 Classify with Nearby ExamplesExamine distance, neighbor count, and voting in k-NN.
9. 09 Look Beyond AccuracySplit data properly and measure the kinds of mistakes that matter.
10. 10 Connect Missing Values to EvaluationPlan a small classification task from preparation through evaluation.