Description
Neural-network models can be built and experimented with inside Octave. This helps students and researchers explore older machine-learning workflows and simple network structures.
It is a machine-learning extension. Model results depend on data quality, training choices, and validation, and should not be treated as automatically reliable.