Mehmet EmrahKonyaTR
← Back to glossary

Neural network

Yapay sinir ağı

D2

A neural network is a model composed of layers of connected computational units and nonlinear operations.

Review status: 2026-11-26

Technical explanation

In a feedforward network, layers apply successive computations, and training adjusts parameters so the network can approximate a target mapping for a task.

Conceptual boundaries

A neural network is a broad model family, whereas deep learning concerns networks whose layered composition gives the model depth.

Provider-neutral example

A team can train a network to map image pixels to a proposed defect category for an inspector to review.

Limitations

Network depth is a structural description and does not by itself establish task performance.

Related concepts

Atomic claims and evidence

  1. 1.1Goodfellow, Bengio, and Courville, Deep Learning, Chapter 6
    Source
    Goodfellow, Bengio, and Courville, Deep Learning, Chapter 6
    Source role
    Authoritative source
    Exact locator
    Chapter 6, p. 164: network composition and directed acyclic graph
    Supported claim
    Feedforward neural networks are commonly represented by composing functions in a directed acyclic graph that maps inputs to outputs.
    Last verification
    Review due
    Scope limitation
    This claim describes feedforward networks and does not define every neural-network architecture.
  2. 2.1Goodfellow, Bengio, and Courville, Deep Learning, Chapter 6
    Source
    Goodfellow, Bengio, and Courville, Deep Learning, Chapter 6
    Source role
    Authoritative source
    Exact locator
    Chapter 6, p. 165: layers and depth of a model
    Supported claim
    In a chain-structured neural network, the functions are called layers and the chain length gives the model its depth.
    Last verification
    Review due
    Scope limitation
    Depth is a structural description and does not by itself establish task performance.
  3. 3.1NIST SP 1500-29, Artificial Intelligence in the Fire Service: Considerations for Implementing Artificial Intelligence into Electronic Safety Equipment
    Source
    NIST SP 1500-29, Artificial Intelligence in the Fire Service: Considerations for Implementing Artificial Intelligence into Electronic Safety Equipment
    Source role
    Authoritative source
    Exact locator
    Section 1.1, p. 5: Neural networks
    Supported claim
    NIST describes a neural net as a model made of layers of connected computational units and nonlinear operations.
    Last verification
    Review due
    Scope limitation
    This is a broad structural definition and does not specify a particular depth, architecture, or task performance.