Unified framework for representing and ranking

Jim Jing Yan Wang, Halima Bensmail

Research output: Contribution to journalArticle

Abstract

In the database retrieval and nearest neighbor classification tasks, the two basic problems are to represent the query and database objects, and to learn the ranking scores of the database objects to the query. Many studies have been conducted for the representation learning and the ranking score learning problems, however, they are always learned independently from each other. In this paper, we argue that there are some inner relationships between the representation and ranking of database objects, and try to investigate their relationships by learning them in a unified way. To this end, we proposed the Unified framework for Representation and Ranking (UR2) of objects for the database retrieval and nearest neighbor classification tasks. The learning of representation parameter and the ranking scores are modeled within one single unified objective function. The objective function is optimized alternately with regard to representation parameter and the ranking scores. Based on the optimization results, iterative algorithms are developed to learn the representation parameter and the ranking scores on a unified way. Moreover, with two different formulas of representation (feature selection and subspace learning), we give two versions of UR2. The proposed algorithms are tested on two challenging tasks - MRI image based brain tumor retrieval and nearest neighbor classification based protein identification. The experiments show the advantage of the proposed unified framework over the state-of-the-art independent representation and ranking methods.

Original languageEnglish
Pages (from-to)2293-2300
Number of pages8
JournalPattern Recognition
Volume47
Issue number6
DOIs
Publication statusPublished - 1 Jun 2014

Fingerprint

Magnetic resonance imaging
Feature extraction
Tumors
Brain
Proteins
Object-oriented databases
Experiments

Keywords

  • Data representation
  • Database retrieval
  • Nearest neighbor classification
  • Ranking score learning

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Signal Processing

Cite this

Unified framework for representing and ranking. / Wang, Jim Jing Yan; Bensmail, Halima.

In: Pattern Recognition, Vol. 47, No. 6, 01.06.2014, p. 2293-2300.

Research output: Contribution to journalArticle

Wang, Jim Jing Yan ; Bensmail, Halima. / Unified framework for representing and ranking. In: Pattern Recognition. 2014 ; Vol. 47, No. 6. pp. 2293-2300.
@article{a0c2f1befa0141008f272c88a6c7d7eb,
title = "Unified framework for representing and ranking",
abstract = "In the database retrieval and nearest neighbor classification tasks, the two basic problems are to represent the query and database objects, and to learn the ranking scores of the database objects to the query. Many studies have been conducted for the representation learning and the ranking score learning problems, however, they are always learned independently from each other. In this paper, we argue that there are some inner relationships between the representation and ranking of database objects, and try to investigate their relationships by learning them in a unified way. To this end, we proposed the Unified framework for Representation and Ranking (UR2) of objects for the database retrieval and nearest neighbor classification tasks. The learning of representation parameter and the ranking scores are modeled within one single unified objective function. The objective function is optimized alternately with regard to representation parameter and the ranking scores. Based on the optimization results, iterative algorithms are developed to learn the representation parameter and the ranking scores on a unified way. Moreover, with two different formulas of representation (feature selection and subspace learning), we give two versions of UR2. The proposed algorithms are tested on two challenging tasks - MRI image based brain tumor retrieval and nearest neighbor classification based protein identification. The experiments show the advantage of the proposed unified framework over the state-of-the-art independent representation and ranking methods.",
keywords = "Data representation, Database retrieval, Nearest neighbor classification, Ranking score learning",
author = "Wang, {Jim Jing Yan} and Halima Bensmail",
year = "2014",
month = "6",
day = "1",
doi = "10.1016/j.patcog.2013.12.003",
language = "English",
volume = "47",
pages = "2293--2300",
journal = "Pattern Recognition",
issn = "0031-3203",
publisher = "Elsevier Limited",
number = "6",

}

TY - JOUR

T1 - Unified framework for representing and ranking

AU - Wang, Jim Jing Yan

AU - Bensmail, Halima

PY - 2014/6/1

Y1 - 2014/6/1

N2 - In the database retrieval and nearest neighbor classification tasks, the two basic problems are to represent the query and database objects, and to learn the ranking scores of the database objects to the query. Many studies have been conducted for the representation learning and the ranking score learning problems, however, they are always learned independently from each other. In this paper, we argue that there are some inner relationships between the representation and ranking of database objects, and try to investigate their relationships by learning them in a unified way. To this end, we proposed the Unified framework for Representation and Ranking (UR2) of objects for the database retrieval and nearest neighbor classification tasks. The learning of representation parameter and the ranking scores are modeled within one single unified objective function. The objective function is optimized alternately with regard to representation parameter and the ranking scores. Based on the optimization results, iterative algorithms are developed to learn the representation parameter and the ranking scores on a unified way. Moreover, with two different formulas of representation (feature selection and subspace learning), we give two versions of UR2. The proposed algorithms are tested on two challenging tasks - MRI image based brain tumor retrieval and nearest neighbor classification based protein identification. The experiments show the advantage of the proposed unified framework over the state-of-the-art independent representation and ranking methods.

AB - In the database retrieval and nearest neighbor classification tasks, the two basic problems are to represent the query and database objects, and to learn the ranking scores of the database objects to the query. Many studies have been conducted for the representation learning and the ranking score learning problems, however, they are always learned independently from each other. In this paper, we argue that there are some inner relationships between the representation and ranking of database objects, and try to investigate their relationships by learning them in a unified way. To this end, we proposed the Unified framework for Representation and Ranking (UR2) of objects for the database retrieval and nearest neighbor classification tasks. The learning of representation parameter and the ranking scores are modeled within one single unified objective function. The objective function is optimized alternately with regard to representation parameter and the ranking scores. Based on the optimization results, iterative algorithms are developed to learn the representation parameter and the ranking scores on a unified way. Moreover, with two different formulas of representation (feature selection and subspace learning), we give two versions of UR2. The proposed algorithms are tested on two challenging tasks - MRI image based brain tumor retrieval and nearest neighbor classification based protein identification. The experiments show the advantage of the proposed unified framework over the state-of-the-art independent representation and ranking methods.

KW - Data representation

KW - Database retrieval

KW - Nearest neighbor classification

KW - Ranking score learning

UR - http://www.scopus.com/inward/record.url?scp=84894305128&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84894305128&partnerID=8YFLogxK

U2 - 10.1016/j.patcog.2013.12.003

DO - 10.1016/j.patcog.2013.12.003

M3 - Article

VL - 47

SP - 2293

EP - 2300

JO - Pattern Recognition

JF - Pattern Recognition

SN - 0031-3203

IS - 6

ER -