The proposed bachelor thesis creates a new, artificial neural network and deep learning based processing chain for electroencephalographic data classification and compares it to a classical processing chain (baseline). This represents a strong interdisciplinary approach, using a biologically inspired deep learning method on a neuroscientific problem while practical implementation requires knowledge in computer science. Spatial and temporal filtering shall be achieved through convolutional layers, as they have brought breakthroughs in various fields of machine learning and pattern recognition in the recent past. For a proof of concept P300 data will be used for training and testing the new node within the pySPACE framework. Result of the Thesis should be an optimized, standard processing chain with high classification accuracy for current and future EEG aided projects to improve human-machine interaction.