Classical ML

12 pages in this section.

Why scikit-learn's fit/predict/transform contract, the train/test/validation split, and the bias-variance trade-off are really one underlying idea - the mental model behind every other page in this section.

Explainer Python Classical Ml Bias Variance

Machine Learning Basics

The ML workflow and core terminology.

Python Classical Ml Basics Machine

Datasets & Splits

Train/test/validation and cross-validation.

Python Classical Ml Datasets Splits

Feature Engineering

Encoding, scaling, and `ColumnTransformer`.

Python Classical Ml Feature Engineering Preprocessing

Pipelines

Composing preprocessing and models leak-free.

Python Classical Ml Pipelines Sklearn

Supervised Learning

Regression and classification algorithms.

Python Classical Ml Supervised Regression

Model Evaluation

Python Classical Ml Evaluation Metrics

Hyperparameter Tuning

Python Classical Ml Hyperparameters Optuna

Unsupervised Learning

Python Classical Ml Unsupervised Clustering

Imbalanced Data

Python Classical Ml Imbalanced Resampling

Gradient Boosting

Python Classical Ml Xgboost Lightgbm

Machine Learning Best Practices

Avoiding leakage and building reproducible models.

Python Classical Ml Best Practices List

Classical ML

Search Documentation

Classical ML