Deep learning is a subfield of machine learning that relies on deep neural network architectures to model highly complex nonlinear relationships.
These models consist of multiple layers that learn hierarchical representations of data, especially in domains such as images, audio, and natural language.
Training is typically performed using optimization techniques such as gradient descent and backpropagation, often supported by large datasets and specialized hardware like GPUs.
Deep learning is a key technology in modern AI systems and underpins applications such as language models, computer vision, and autonomous systems.
Definition:
Deep learning is a machine learning approach that uses multi-layer neural networks to learn complex patterns and representations from large datasets.
Source:
Goodfellow, Bengio, Courville – Deep Learning, MIT Press.