Learning in neural networks with local minima

An attempt is made to study learning in neural networks with local minima. For small learning parameters {η}, the transition time from one mimimum to another is asymptotically given by exp({η} ̃ \~{}{}/{η}), with {η} ̃ \~{}{}, a constant independent of {η}, called the reference learning parameter. A general scheme to calculate the reference learning parameter is presented. This scheme is valid for a large class of learning rules.