Multiple instance classification via quadratic programming

Emel Şeyma Küçükaşcı (), Mustafa Gökçe Baydoğan () and Z. Caner Taşkın ()
Additional contact information
Emel Şeyma Küçükaşcı: Istanbul Commerce University
Mustafa Gökçe Baydoğan: Boğaziçi University
Z. Caner Taşkın: Boğaziçi University

Journal of Global Optimization, 2022, vol. 83, issue 4, No 1, 639-670

Abstract: Abstract Multiple instance learning (MIL) is a variation of supervised learning, where data consists of labeled bags and each bag contains a set of instances. Unlike traditional supervised learning, labels are not known for the instances in MIL. Existing approaches in the literature make use of certain assumptions regarding the instance labels and propose mixed integer quadratic programs, which introduce computational difficulties. In this study, we present a novel quadratic programming (QP)-based approach to classify bags. Solution of our QP formulation links the instance-level contributions to the bag label estimates, and provides a linear bag classifier along with a decision threshold. Our approach imposes no additional constraints on relating instance labels to bag labels and can be adapted to learning applications with different MIL assumptions. Unlike existing specialized heuristic approaches to solve previous MIL formulations, our QP models can be directly solved to optimality using any commercial QP solver. Also, kindly confirm Our computational experiments show that proposed QP formulation is efficient in terms of solution time, overcoming a main drawback of previous optimization algorithms for MIL. We demonstrate the classification success of our approach compared to the state-of-the-art methods on a wide range of real world datasets.

Keywords: Multiple instance learning; Classification; Quadratic programming (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s10898-021-01120-0 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:jglopt:v:83:y:2022:i:4:d:10.1007_s10898-021-01120-0

Ordering information: This journal article can be ordered from
http://www.springer. ... search/journal/10898

DOI: 10.1007/s10898-021-01120-0

Access Statistics for this article

Journal of Global Optimization is currently edited by Sergiy Butenko

More articles in Journal of Global Optimization from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().