Teknosys learning library

Recursive networks and RNTN

Recursive neural networks combine representations along a structure such as a sentence parse tree, rather than simply following token order. Recursive neural tensor networks (RNTNs) add tensor-based interactions to model how words and phrases compose. These architectures are associated with tasks such as phrase-level sentiment analysis, where negation and word combinations matter. Recursive and recurrent networks are distinct, even though both are sometimes abbreviated informally.

Educational material. Verify examples against current documentation and applicable licenses. Report a correction.