A Precept - Driven Learning Algorithm

Machine learning is an attempt at devising mechanismsthat machines can use to learn, rather than being explicitlyprogrammed for, real-world applications. The goal oflearning systems is to generalize. Generalization is basedon the set of critical features the system has available.Training set learners typically extract critical featuresfrom a random set of examples drawn from experimentation.This approach can beneficially be extended by endowing thesystem with some a priori knowledge, in the form of precepts.Advantages of the augmented system include speed-up, improvedgeneralization and greater parsimony. This thesis presents a precept-driven learning algorithm.The main characteristics of the algorithm include: 1)neurally inspired architecture, 2) bounded learning andexecution times, and 3) ability to handle both correct andincorrect precepts. Results of simulations on real-worlddata demonstrate promise. COMMITTEE APPROVAL: ___________________________________Tony Martinez, Committee Chair ___________________________________David Embley, Committee Member ___________________________________David Embley, Graduate Coordinator