The similarity between artificial and biological neural networks is only superficial.

Evidence

Artificial neural networks (ANNs) (Haykin, 1994; Kohonen, 1990; Lippmann, 1987; Luo and Unbehauen, 1997) are signal and information processing techniques that learn or are trained to acquire a targeted behaviour, typically an input–output relationship of interest. Some principles of ANNs are found by trying to simulate the behaviour of biological neural nets, but the similarity is not very strong, and true neural networks, for example in the brain, may work in quite different ways. — Ville Pulkki, Matti Karjalainen, Communication Acoustics- An…, 450

Categorization

• DEFINE: Defines ANNs and notes the deliberate caveat that their resemblance to brain networks is not strong.
• XREF: Relevant to the reader’s acoustic-communication application focus: ANN-based speech recognition figures heavily in the book’s audio chapters.