Bidirectional associative memory (BAM)‎ neural network for tries structures searching

Abstract EN

In this work, the relationship between the searching in data structures and neural network is presented.

Bidirectional Associative Memory (BAM)neural network is used for searching any word in trie structures method with Hamming distance by introducing weighting matrix into the connection matrix.

This generalization is validated to increase storage capacity, the design algorithm for BAM net is to search word trie structure have 0 ((P ? (n + m)) computational complexity, where p is the number of prototype pairs and n,m are the dimensions of input / output bipolar vectors.