Compact coding for hyperplane classifiers in heterogeneous environment

Hao Shao, Bin Tong, Einoshin Suzuki

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

Transfer learning techniques have witnessed a significant development in real applications where the knowledge from previous tasks are required to reduce the high cost of inquiring the labeled information for the target task. However, how to avoid negative transfer which happens due to different distributions of tasks in heterogeneous environment is still a open problem. In order to handle this kind of issue, we propose a Compact Coding method for Hyperplane Classifiers (CCHC) under a two-level framework in inductive transfer learning setting. Unlike traditional methods, we measure the similarities among tasks from the macro level perspective through minimum encoding. Particularly speaking, the degree of the similarity is represented by the relevant code length of the class boundary of each source task with respect to the target task. In addition, informative parts of the source tasks are adaptively selected in the micro level viewpoint to make the choice of the specific source task more accurate. Extensive experiments show the effectiveness of our algorithm in terms of the classification accuracy in both UCI and text data sets.

Original languageEnglish
Title of host publicationMachine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2011, Proceedings
Pages207-222
Number of pages16
EditionPART 3
DOIs
Publication statusPublished - Sep 9 2011
EventEuropean Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML PKDD 2011 - Athens, Greece
Duration: Sep 5 2011Sep 9 2011

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 3
Volume6913 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

OtherEuropean Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML PKDD 2011
CountryGreece
CityAthens
Period9/5/119/9/11

Fingerprint

Heterogeneous Environment
Hyperplane
Transfer Learning
Macros
Classifiers
Coding
Classifier
Inductive Learning
Costs
Target
Experiments
Open Problems
Encoding
Experiment
Similarity

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Shao, H., Tong, B., & Suzuki, E. (2011). Compact coding for hyperplane classifiers in heterogeneous environment. In Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2011, Proceedings (PART 3 ed., pp. 207-222). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6913 LNAI, No. PART 3). https://doi.org/10.1007/978-3-642-23808-6_14

Compact coding for hyperplane classifiers in heterogeneous environment. / Shao, Hao; Tong, Bin; Suzuki, Einoshin.

Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2011, Proceedings. PART 3. ed. 2011. p. 207-222 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6913 LNAI, No. PART 3).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Shao, H, Tong, B & Suzuki, E 2011, Compact coding for hyperplane classifiers in heterogeneous environment. in Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2011, Proceedings. PART 3 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 3, vol. 6913 LNAI, pp. 207-222, European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML PKDD 2011, Athens, Greece, 9/5/11. https://doi.org/10.1007/978-3-642-23808-6_14
Shao H, Tong B, Suzuki E. Compact coding for hyperplane classifiers in heterogeneous environment. In Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2011, Proceedings. PART 3 ed. 2011. p. 207-222. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3). https://doi.org/10.1007/978-3-642-23808-6_14
Shao, Hao ; Tong, Bin ; Suzuki, Einoshin. / Compact coding for hyperplane classifiers in heterogeneous environment. Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2011, Proceedings. PART 3. ed. 2011. pp. 207-222 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 3).
@inproceedings{dd16dfea20214c1eb81fb9bd29d990e5,
title = "Compact coding for hyperplane classifiers in heterogeneous environment",
abstract = "Transfer learning techniques have witnessed a significant development in real applications where the knowledge from previous tasks are required to reduce the high cost of inquiring the labeled information for the target task. However, how to avoid negative transfer which happens due to different distributions of tasks in heterogeneous environment is still a open problem. In order to handle this kind of issue, we propose a Compact Coding method for Hyperplane Classifiers (CCHC) under a two-level framework in inductive transfer learning setting. Unlike traditional methods, we measure the similarities among tasks from the macro level perspective through minimum encoding. Particularly speaking, the degree of the similarity is represented by the relevant code length of the class boundary of each source task with respect to the target task. In addition, informative parts of the source tasks are adaptively selected in the micro level viewpoint to make the choice of the specific source task more accurate. Extensive experiments show the effectiveness of our algorithm in terms of the classification accuracy in both UCI and text data sets.",
author = "Hao Shao and Bin Tong and Einoshin Suzuki",
year = "2011",
month = "9",
day = "9",
doi = "10.1007/978-3-642-23808-6_14",
language = "English",
isbn = "9783642238079",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 3",
pages = "207--222",
booktitle = "Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2011, Proceedings",
edition = "PART 3",

}

TY - GEN

T1 - Compact coding for hyperplane classifiers in heterogeneous environment

AU - Shao, Hao

AU - Tong, Bin

AU - Suzuki, Einoshin

PY - 2011/9/9

Y1 - 2011/9/9

N2 - Transfer learning techniques have witnessed a significant development in real applications where the knowledge from previous tasks are required to reduce the high cost of inquiring the labeled information for the target task. However, how to avoid negative transfer which happens due to different distributions of tasks in heterogeneous environment is still a open problem. In order to handle this kind of issue, we propose a Compact Coding method for Hyperplane Classifiers (CCHC) under a two-level framework in inductive transfer learning setting. Unlike traditional methods, we measure the similarities among tasks from the macro level perspective through minimum encoding. Particularly speaking, the degree of the similarity is represented by the relevant code length of the class boundary of each source task with respect to the target task. In addition, informative parts of the source tasks are adaptively selected in the micro level viewpoint to make the choice of the specific source task more accurate. Extensive experiments show the effectiveness of our algorithm in terms of the classification accuracy in both UCI and text data sets.

AB - Transfer learning techniques have witnessed a significant development in real applications where the knowledge from previous tasks are required to reduce the high cost of inquiring the labeled information for the target task. However, how to avoid negative transfer which happens due to different distributions of tasks in heterogeneous environment is still a open problem. In order to handle this kind of issue, we propose a Compact Coding method for Hyperplane Classifiers (CCHC) under a two-level framework in inductive transfer learning setting. Unlike traditional methods, we measure the similarities among tasks from the macro level perspective through minimum encoding. Particularly speaking, the degree of the similarity is represented by the relevant code length of the class boundary of each source task with respect to the target task. In addition, informative parts of the source tasks are adaptively selected in the micro level viewpoint to make the choice of the specific source task more accurate. Extensive experiments show the effectiveness of our algorithm in terms of the classification accuracy in both UCI and text data sets.

UR - http://www.scopus.com/inward/record.url?scp=80052400362&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80052400362&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-23808-6_14

DO - 10.1007/978-3-642-23808-6_14

M3 - Conference contribution

SN - 9783642238079

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 207

EP - 222

BT - Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2011, Proceedings

ER -