Computer Science and Engineering, Department of


Date of this Version



University of Nebraska-Lincoln, Computer Science and Engineering
Technical Report # TR-UNL-CSE-2003-0008


In this paper we prove a new upper bound on learning monotone DNF formulae in an exact learning model called the teaching assistant model. This model, introduced in one of our earlier works, uses teaching assistant classes for classifying the complexity of learning problems. Teaching assistant classes are a learning-theoretic analog of complexity classes. We show that monotone DNF are learnable using a teaching assistant in the class SPP. On the other hand, concept classes such as k-CNF or Boolean circuits are not learnable using an SPP teaching assistant unless Np SPP. We also observe that the recent SPP algorithm for Graph Isomophism problem due to Arvind and Kurur can be adapted to learn the concept of permutation groups using an SPP teaching assistant. This solves an open problem for our earlier work.