A neural network for tornado diagnosis: Managing local minima

There exist radar-based algorithms designed to detect circulations in the a tmosphere. Not all detected circulations, however, are associated with torn ados on the ground. Outlined herein is the development of a multi-layered p erceptron designed to classify the two types of circulations - nontornadic and tornadic - based on various attributes of the circulations. Special emp hasis is placed on the role of local minima in determining the optimal arch itecture via bootstrapping, and on the performance of the network in terms of probabilistic measures.