1 AIT Asian Institute of Technology

Neural networks for logic programming

AuthorLe Phung Long
Call NumberAIT Thesis no.CS-94-30
Subject(s)Neural networks (Computer science)
NoteA thesis submitted in partial fulfillment of the requirements for the degree of Master of Science, School of Engineering and technology
PublisherAsian Institute of Technology
AbstractThe thesis is divided into two parts . The first part proposes a Hopfield network to find the Well-Founded Model of a given logic program. The definition of Well-Founded model is converted to a set of propositions. The Hopfield network is described to represent the model of these propositions, and Well-Founded Model of logic program is found when the network is stable. The second part presents an idea to deal with the problem of variable bindings m connectionist models. The 3-layer feed-forward neural networks for logic program are proposed by Holldobler et al. for propositional logic program. We propose an extended model for datalog program. A generalization of the notions of inputs and outputs of neural nodes is the main idea. In the conventional connectionist model, inputs and outputs conveyed on the connections among nodes are scalar values, whilst in our model , the instantiation matrices of instances of predicate are conveyed. By this way, all instances of predicate are possible to be represented in one time.
Year1994
TypeThesis
SchoolSchool of Engineering and Technology (SET)
DepartmentOther Field of Studies (No Department)
Academic Program/FoSComputer Science (CS)
Chairperson(s)Phan Minh Dung
Examination Committee(s)Huynh Ngoc Phien ;Sadananda, Ramakoti
Scholarship Donor(s)Thai AID Grant
Degree Thesis (M.Sc.) - Asian Institute of Technology, 1994


Usage Metrics
View Detail0
Read PDF0
Download PDF0