Scikit-Learn Tutorial with Examples.
Scikit-Learn tutorial for machine learning: estimators, pipelines, cross-validation, feature engineering and model selection.
Beginner
Start here — no prior experience needed
Scikit-Learn Learning Roadmap
A structured path from sklearn beginner to ML practitioner — from your first classifier to building production-ready pipelines.
Introduction to Scikit-Learn
Learn scikit-learn's fit/predict/score estimator API by training a first classifier and regressor, with the real output of every example shown.
Scikit-Learn Feature Engineering
Scale, encode, impute and create features in scikit-learn, with real runs showing when scaling matters and how default encodings silently go wrong.
Scikit-Learn Projects
End-to-end ML projects that take you from raw data to deployed models — covering classification, regression, clustering, and production pipelines.
Intermediate
For developers with core concepts down
Scikit-Learn Model Evaluation
Evaluate scikit-learn models with cross-validation, precision and recall, ROC curves, regression metrics and calibration, with real output for each.
Scikit-Learn Pipelines
Chain preprocessing and models into scikit-learn Pipelines and ColumnTransformers, with a real run showing how preprocessing outside one leaks data.
Scikit-Learn Classification
Train and compare logistic regression, random forests, gradient boosting and SVMs in scikit-learn, then tune the decision threshold without leaking.
Scikit-Learn Hyperparameter Tuning
Tune models with GridSearchCV, RandomizedSearchCV, successive halving and Optuna, and measure how optimistic a tuned model's best CV score really is.
Scikit-Learn Regression
Train linear, Ridge, Lasso and gradient-boosted regressors in scikit-learn, with real runs of regularisation, prediction intervals and residual checks.
Scikit-Learn Clustering
Group unlabeled data with K-Means, DBSCAN and agglomerative clustering in scikit-learn, and see with real runs where each one fails and why.
Advanced
Production-grade patterns for experienced engineers
XGBoost and LightGBM
Train XGBoost and LightGBM with early stopping and native categoricals, compare them to scikit-learn's own booster, and fix old XGBoost tutorial code.
Scikit-Learn Advanced Model Selection
Use validation and learning curves, group-aware and time-series splits, and out-of-fold predictions in scikit-learn, with real runs of each leakage trap.