Multimodal acquisition from several kinds of sensors has enable the exploitation of information that may be of a complementary nature. An adequate analysis of the data available together could improve the results of individual analyses for each of the modes of the data acquisition. In this paper, we propose a procedure to classify a publicly available dataset of functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) recordings measured simultaneously. The proposed method is based on early fusion at the level of the features extracted from data and late fusion at the level of the results from several classifiers. Besides, spatial and time synchronization is approached due to the differences in sampling frequency of the two modalities. This later is done using an advanced fusion method called alpha integration, which is compared with the results of the individual classifiers and the fusion using the mean. The results show improvement of classification accuracy and F1 index obtained by the fusion.