A local minimum is defined as a point such that all points in a neighbourhood have an error value greater than or equal to the error value in that point. It is�...
The main result of this paper is that all stationary points of the 2-2-1 XOR network with finite weights cannot be local minima: they are absolute minima with�...
It is proved that the error surface of the two-layer XOR network with two hidden units has a number of regions with local minima such that one or both�...
A local minimum is defined as a point such that all points in a neighbourhood have an error value greater than or equal to the error value in that point. It is�...
We investigated the error surface of the XOR problem with a 2-2-1 network with sigmoid transfer functions. In this paper it is proved that.
So, for finite weights no local minima occur. The proof results from a careful analysis of the Taylor series expansion around the stationary points. For some�...
It is proved that the error surface of the two-layer XOR network with two hidden units has a number of regions with local minima such that one or both�...
Recommendations � The local minima of the error surface of the 2-2-1 XOR network. All local minima of the error surface of the 2�2�1 XOR network are described.
This counterexample consists of a region of local minima with nonzero error on the error surface of a neural network with three hidden nodes when presented with�...
Nov 8, 2011 � Yes, neural networks can get stuck in local minima, depending on the error surface. However this abstract suggests that there are no local�...