In this paper, we study a learning procedure from positive data for bounded unions of certain class of languages. Our key tools are the notion of characteristic sets and hypergraphs. We generate hy-pergraphs from given positive data and exploit them in order to find characteristic sets.
展开▼