Neural Network

NOOR-uhl NET-wurk /ˈnjʊərəl ˈnɛtwɜːrk/
nounC1GRETOEFLIELTS
formaltechnical

A computing system inspired by the structure and function of the human brain, composed of interconnected nodes (neurons) organized in layers, that can learn from data.

Neural Networks are a core component of Machine Learning, particularly Deep Learning. They excel at tasks like pattern recognition and classification by adjusting the 'weights' of connections between their artificial neurons during training.

The story behind the word. The term 'Neural Network' draws directly from biology. 'Neural' comes from the Greek `neuron`, meaning 'nerve,' referring to the nervous system. 'Network' comes from Old English `net` and `weorc` (work), referring to an interconnected system. The term was chosen to reflect its inspiration from the interconnected neurons and synapses of the biological brain, aiming to mimic its learning capabilities.

Word relationships

  • artificial neural network: More explicit, to distinguish from biological neural networks.

Commonly confused with

  • Deep Learning: Deep Learning is a type of Machine Learning that uses Neural Networks with many layers ('deep').

Word family

  • neural (adjective): Relating to a nerve or the nervous system.

Collocations

  • deep Neural Network
  • train a Neural Network
  • Neural Network architecture
  • recurrent Neural Network

Example sentences

  • "The image recognition software uses a deep Neural Network to identify objects within photographs." Explaining how image recognition works.
  • "Building an effective Neural Network requires careful design of its layers and activation functions." Discussing the architecture of an AI system.

Memory hook

Think of a 'network' of 'neurons' (like brain cells) working together in a computer. Brain-like Network

When not to use

Do not use 'Neural Network' as a synonym for 'Artificial Intelligence' or 'Machine Learning.' It is a specific type of model or architecture within those fields.

Fun facts

  • The first conceptual model of an artificial neuron, the McCulloch-Pitts model, was proposed in 1943.
  • Early neural networks faced a 'winter' of funding and interest due to limitations, before resurging with increased computing power and data.
LangBlitz
Take your vocabulary further with LangBlitz

Our English vocabulary app: FSRS spaced repetition, 5,000+ curated words across 119 topic groups, CEFR A1 to C2. Explore your mastery with the beautiful Vocabulary World feature.