This project explored how decision trees and neural networks can be used to identify radar signatures associated with severe and non-severe weather. Our task was to design a data mining technique which identifies three classes of severe thunderstorms and a non-severe class from a set of training data and test data comprised of different events with the same attributes for each event. The training data also contained know classifications. Decision trees have long been known to be efficient classifiers by pruning away branches that do not contribute any information gain. The Waikato Environment for Knowledge Analysis (WEKA) tool was used to develop a decision tree which pared away attributes from the original training file that do not contribute much to information gain for the given classification set. A decision tree can also be used for classification but it was felt the type of classification problem contained may non-linear relationships that a decision tree might not be able to identify. However, a neural network is capable of identifying non-linear relationships in data. The resulting subset of attributes from the decision tree was used as input to a neural network which attempts to identify these non-linear relationships to produce a better classification algorithm.