Dictionary of Applied Machine Learning

Supported by Forward-Looking AI Governance in Banking and Insurance (FLAIG) (Business Finland, record no. 7360/31/2023); A Mathematical Theory of Trustworthy Federated Learning (Research Council of Finland, decision no. 363624); A Mathematical Theory of Federated Learning (TRUST-FELT) (Jane and Aatos Erkko Foundation, decision no. A835).

Course edition, CC BY 4.0. Each term links a typeset PDF and a Python script that recomputes what the entry states.

How this book is built · List of symbols · For instructors · The linters · People

• accuracy
• artificial neural network
• attention
• autoencoder
• baseline
• boosting
• bootstrap aggregating
• classification
• clustering
• concept activation vector
• confusion matrix
• convex
• covariance matrix
• data
• data leakage
• data point
• dataset
• decision tree
• dimensionality reduction
• distribution shift
• eigenvalue
• eigenvalue decomposition
• empirical risk minimization
• ensemble
• expectation–maximization
• explainability
• explainable artificial intelligence
• explanation
• feature
• feature importance
• feature learning
• fine-tuning
• Gaussian mixture model
• Gaussian random variable
• generalization
• gradient
• gradient descent
• Hilbert space
• hyperparameter
• hypothesis
• hypothesis space
• independent and identically distributed assumption
• inner product
• interpretability
• interpretable machine learning
• kernel method
• kernel ridge regression
• k-means
• label
• large language model
• learning rate
• linear regression
• logistic regression
• loss
• machine learning pipeline
• matrix
• mean
• model
• norm
• online learning
• overfitting
• policy
• pretraining
• principal component analysis
• probability distribution
• projection
• random forest
• random variable
• regularization
• reinforcement learning
• reward
• self-supervised learning
• singular value decomposition
• Sobolev space
• stochastic gradient descent
• supervised learning
• support vector machine
• tabular data
• test set
• training set
• transfer learning
• transformer
• transparency
• unsupervised learning
• validation set
• variance