An estimation framework in a forward-propagation learning rule
Yoshihiro Ohama, Naohiro Fukumura, Yasuo UNO · Society of Instrument and Control Engineers of Japan · 2004
A forward-propagation learning rule has been proposed to acquire neural inverse models. This rule can solve a credit assignment problem based on Newton-like method. In the current work, we discuss how to estimate the parameters of a multi-layered neural network based on the credit assignment. The suitability of the proposed estimation framework is confirmed by computer simulation.