Machine learning foundations with scikit-learn

Machine learning is a way of building programs that learn patterns from examples instead of following hand written rules. It powers spam filters, demand forecasts, fraud alerts and recommendation systems. Before deep learning or large language models, every practitioner needs the foundations: how models learn, how to measure them honestly and how to avoid the mistakes that make results look better than they are.

This course teaches those foundations with scikit-learn, the standard Python library for classical machine learning. We work on two realistic projects: predicting whether a customer will cancel a subscription, and estimating house prices.

What you will learn

• The machine learning workflow from question to model
• Train, validation and test splits, and why leakage ruins results
• Linear and logistic regression, decision trees, random forests and gradient boosting
• Preprocessing with pipelines and column transformers
• Metrics for classification and regression, and choosing the right one
• Cross validation and hyperparameter tuning
• Handling imbalanced classes, interpreting models and saving them for later use

Approach

We start every project with a simple baseline, then add complexity only when it improves validated results. You will learn to distrust a single great number and to ask how it was measured. Code is short and explained line by line.

You need basic Python and pandas. The maths is explained intuitively; school algebra is enough.