Modelling with constructive backpropagation

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

AdaNet: Adaptive Structural Learning of Artificial Neural Networks (2016, background): “There has been extensive work involving structure learning for neural networks (e.g. [Kwok and Yeung, 1997, Leung et al., 2003, Islam et al., 2003, Lehtokangas, 1999, Islam et al., 2009]).”
A structure optimisation algorithm for feedforward neural network construction (2013, background): “. n), isibox@sina.com structure (GCS) [7], constructive back-propagation (CBP) [8], adaptively constructing multilayer FNN [9], extreme learning machine with adaptive growth of hidden nodes (AG-ELM) [10], enhanced incremental extreme learning machine (EI-ELM) [11], and repair RBF neural network (RRB”
An integrated growing-pruning method for feedforward network training (2008, methodology): “Also it is shown in [19] that adding multiple hidden units at a time and training together is much more efficient than adapting many single units trained independently.”