Back to glossary

Training

Eğitim

D1

Training is the process of improving a system’s performance with machine-learning techniques, commonly by fitting learnable model parameters using training data.

Review status: 2027-02-23

Technical explanation

In a machine-learning workflow, training produces a model from training data and can improve a system’s performance.

Conceptual boundaries

Training is the learning process, whereas inference uses a model to generate outputs; it is not the same as merely collecting data or putting a model into use.

Provider-neutral example

A team can fit a model’s learnable parameters on historical demand data before using it to propose next-week quantities.

Limitations

This entry concerns machine-learning training and does not prescribe a particular training method.

Related concepts

Atomic claims and evidence

  1. 1.1OECD, Explanatory Memorandum on the Updated OECD Definition of an AI System
    Source
    OECD, Explanatory Memorandum on the Updated OECD Definition of an AI System
    Source role
    Authoritative source
    Exact locator
    Building AI systems and models, machine-learning approaches, p. 7
    Supported claim
    OECD describes training as the process of improving a system’s performance using machine-learning techniques.
    Last verification
    Review due
    Scope limitation
    This description concerns machine-learning training and does not cover every kind of software development.
  2. 2.1Regulation (EU) 2024/1689 (Artificial Intelligence Act)
    Source
    Regulation (EU) 2024/1689 (Artificial Intelligence Act)
    Source role
    Authoritative source
    Exact locator
    Article 3(29)
    Supported claim
    The EU AI Act defines training data as data used to train an AI system through fitting its learnable parameters.
    Last verification
    Review due
    Scope limitation
    This is the Act’s terminology and does not prescribe a particular training method.