Electronic Colloquium on Computational Complexity Bounds for the Computational Power and Learning Complexity of Analog Neural Nets

It is shown that high order feedforward neural nets of constant depth with piecewise polynomial activation functions and arbitrary real weights can be simulated for boolean inputs and outputs by neural nets of a somewhat larger size and depth with heaviside gates and weights from f?1; 0; 1g. This provides the rst known upper bound for the computational power of the former type of neural nets. It is also shown that in the case of rst order nets with piecewise linear activation functions one can replace arbitrary real weights by rational numbers with polynomially many bits, without changing the boolean function that is computed by the neural net. In order to prove these results we introduce two new methods for reducing nonlinear problems about weights in multi-layer neural nets to linear problems for a transformed set of parameters. These transformed parameters can be interpreted as weights in a somewhat larger neural net. As another application of our new proof technique we show that neural nets with piecewise polynomial activation functions and a constant number of analog inputs are probably approximately learnable (in Valiant's model for PAC-learning). subject "MAIL ME CLEAR", body "pub/eccc/ftpmail.txt"