Towards an automatic detection system of signals at cotopaxi volcano

Currently, Cotopaxi volcano has increased its seismic activity. Efficient and reliable tools are required for volcano monitoring, and early and effective emergency alerts are necessary. This article presents a supervised classification-based detection of seismic-volcanic and non-volcanic events recorded during 2010. Our algorithm use four features acquired using energy of approximation and detail coefficients of the wavelet decomposition analyzed with Daubechies and Symlet families. The classification of events was performed by the algorithm of decision trees with embedded techniques: cross-validation and pruning for reduction in the number features. The best results are obtained by application of the mother wavelet Symlet with accuracy of 99% and precision of 98%.