van der Schaar, M., Delory, E., André, M. Comparison of Gaussian based methods to classify sperm whale (Physeter macrocephalus) clicks Proceedings of the Sea Tech Week, Brest, France, Oct 2006 Résumé: Sperm whales often dive in groups for foraging, producing continuous series of clicks that result in mixed signals in the acoustic recordings. When the click trains of individual animals need to be reconstructed, a classi?cation method is required that can assign a click to an individual. This report looks at two classi?cation methods that can use Gaussian functions, a radial basis function (RBF) network, and support vector machines (SVM). The methods were tested on data recordings from 7 individually diving male sperm whales, 6 datasets containing single click trains and the 7th containing data from a complete dive. Both algorithms could distinguish between the clicks of the 7 different whales, butthe SVM seemed to have better generalization towards unknown data. Projet: Noise pollution effects on cetaceans