Seismic event characterization is often accomplished using algorithms based only on information received at seismological stations located closest to the particular event, ignoring historical data received at those stations; these data are stored and unseen at this stage. This characterization process can delay emergency response, costing valuable time that would be useful to mitigate the adverse effects on the affected population. Data from seismological stations are recorded during many events, which have been characterized by classical methods, and can be used as previous ¿knowledge¿ to train such stations for pattern recognition. This knowledge can be used to make faster characterizations using only one three-component broadband station applying bio-inspired algorithms or recently developed stochastic methods, such as kernel methods. We trained a Support Vector Machine (SVM) algorithm with seismograph data recorded by INGEOMINAS¿s National Seismological Network at a three-component station located near Bogota (Colombia). As input model descriptors, we used: (1) the integral of the Fourier transform/power spectrum for each component, divided in 7 windows of 2 seconds and beginning at the P onset time; and (2) the ratio between the calculated logarithm of magnitude (Mb) and epicentral distance. We used 986 events of magnitude higher than 3 recorded from late 2003 to 2008. The algorithm classifies events with magnitude-distance ratios greater than a background value, which is a measure of severity of possible damage caused by an earthquake. Determination of this value allows an estimation of the magnitude, with a known epicentral distance, calculated by the difference between P and S onset times. This rapid (< 20 seconds) magnitude estimate can be used for rapid response strategies. Results obtained in this work confirm that many hypocentral parameters and a rapid location of a seismic event can be obtained using a few seconds of signal registered at a single stati