Papers presented at the Fourth Workshop on Parallel Architectures and Neural Networks, organized by the International Institute for Advanced Scientific Studies, in collaboration with other Italian Institutions.