Learning to Satisfy

TitleLearning to Satisfy
Publication TypeReport
Year of Publication2006
AuthorsCoates, M. J., B. Erikkson, R. D. Nowak, and C. Scott
Date Published11/2006
InstitutionDepartment of Electrical and Computer Engineering, McGill University
CityMontreal, QC, Canada
TypeTechnical Report
Abstract

This paper investigates a class of learning problems called learning satisfiability (LSAT) problems, where the goal is to learn a set in the input (feature) space that satisfies a number of desired output (label/response) properties. LSAT problems are motivated, in part, by applications in computational finance, and an experi- mental investigation of LSAT in the context of portfolio selection is reported. A distinctive aspect of LSAT problems is that the output behavior is assessed only on the solution set, whereas in most statistical learning problems output behavior is evaluated over the entire input space. Consequently, certain learning criteria arising naturally in LSAT problems require a novel large deviation bounding technique.

Refereed DesignationDoes Not Apply
AttachmentSize
coates_TechReport06.pdf184.77 KB