A new strategy for constructing higher order neural networks with multiplication units
D. Li, Koutaro Hirasawa, Jinglu Hu · Society of Instrument and Control Engineers of Japan · 2003
This paper presents a constructive method for higher-order neural networks with multiplication units. The proposed method provides a flexible mechanism for incremental network growth where higher order terms can be naturally generated and combinatorial explosion in conventional higher order units can be avoided. Simulation results on parity problems and a function approximation problem are given to highlight the capability of the proposed network.