HS2: Active Learning over Hypergraphs.

We propose a hypergraph-based active learning scheme which we term HS2; HS2 generalizes the previously reported algorithm S2 originally proposed for graph-based active learning with pointwise queries. Our HS2 method can accommodate hypergraph structures and allows one to ask both pointwise queries and pairwise queries. Based on a novel parametric system particularly designed for hypergraphs, we derive theoretical results on the query complexity of HS2 for the above described generalized settings. Both the theoretical and empirical results show that HS2 requires a significantly fewer number of queries than S2 when one uses S2 over a graph obtained from the corresponding hypergraph via clique expansion.