Ensembles based on Conformal Instance Transfer

Shuang Zhou; Evgueni Smirnov; Gijs Schoenmakers; A Gammerman; V Vovk; Z Luo

Ensembles based on Conformal Instance Transfer

Shuang Zhou^*, Evgueni Smirnov, Gijs Schoenmakers, A Gammerman, V Vovk, Z Luo

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceeding › Academic › peer-review

3 Downloads (Pure)

Abstract

In this paper we propose a new ensemble method based on conformal instance transfer. The method combines feature selection and source-instance selection to avoid negative transfer in a model-independent way. It was tested experimentally for different types of classifiers on several benchmark data sets. The experiment results demonstrate that the new method is capable of outperforming significantly standard instance transfer methods.

Original language	English
Title of host publication	CONFORMAL AND PROBABILISTIC PREDICTION AND APPLICATIONS, VOL 105
Publisher	JMLR - Journal of Machine Learning Research
Number of pages	20
Volume	105
Publication status	Published - 2019
Event	8th Symposium on Conformal and Probabilistic Prediction and Applications - Varna, Bulgaria Duration: 9 Sept 2019 → 11 Sept 2019 Conference number: 8 https://cml.rhul.ac.uk/copa2019/

Publication series

Series	Proceedings of Machine Learning Research
Volume	105
ISSN	2640-3498

Conference

Conference	8th Symposium on Conformal and Probabilistic Prediction and Applications
Abbreviated title	COPA 2019
Country/Territory	Bulgaria
City	Varna
Period	9/09/19 → 11/09/19
Internet address	https://cml.rhul.ac.uk/copa2019/

Keywords

Instance Transfer
Conformal Prediction
Ensembles

Access to Document

zhou19a

Cite this

@inproceedings{6ede7862827247f081662bfc3b2604d0,

title = "Ensembles based on Conformal Instance Transfer",

abstract = "In this paper we propose a new ensemble method based on conformal instance transfer. The method combines feature selection and source-instance selection to avoid negative transfer in a model-independent way. It was tested experimentally for different types of classifiers on several benchmark data sets. The experiment results demonstrate that the new method is capable of outperforming significantly standard instance transfer methods.",

keywords = "Instance Transfer, Conformal Prediction, Ensembles",

author = "Shuang Zhou and Evgueni Smirnov and Gijs Schoenmakers and A Gammerman and V Vovk and Z Luo",

year = "2019",

language = "English",

volume = "105",

series = "Proceedings of Machine Learning Research",

publisher = "JMLR - Journal of Machine Learning Research ",

booktitle = "CONFORMAL AND PROBABILISTIC PREDICTION AND APPLICATIONS, VOL 105",

address = "United States",

note = "8th Symposium on Conformal and Probabilistic Prediction and Applications, COPA 2019 ; Conference date: 09-09-2019 Through 11-09-2019",

url = "https://cml.rhul.ac.uk/copa2019/",

}

Zhou, S, Smirnov, E , Schoenmakers, G, Gammerman, A, Vovk, V & Luo, Z 2019, Ensembles based on Conformal Instance Transfer. in CONFORMAL AND PROBABILISTIC PREDICTION AND APPLICATIONS, VOL 105. vol. 105, JMLR - Journal of Machine Learning Research , Proceedings of Machine Learning Research, vol. 105, 8th Symposium on Conformal and Probabilistic Prediction and Applications, Varna, Bulgaria, 9/09/19.

Ensembles based on Conformal Instance Transfer. / Zhou, Shuang; Smirnov, Evgueni ; Schoenmakers, Gijs et al.
CONFORMAL AND PROBABILISTIC PREDICTION AND APPLICATIONS, VOL 105. Vol. 105 JMLR - Journal of Machine Learning Research , 2019. (Proceedings of Machine Learning Research, Vol. 105).

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceeding › Academic › peer-review

TY - GEN

T1 - Ensembles based on Conformal Instance Transfer

AU - Zhou, Shuang

AU - Smirnov, Evgueni

AU - Schoenmakers, Gijs

AU - Gammerman, A

AU - Vovk, V

AU - Luo, Z

N1 - Conference code: 8

PY - 2019

Y1 - 2019

N2 - In this paper we propose a new ensemble method based on conformal instance transfer. The method combines feature selection and source-instance selection to avoid negative transfer in a model-independent way. It was tested experimentally for different types of classifiers on several benchmark data sets. The experiment results demonstrate that the new method is capable of outperforming significantly standard instance transfer methods.

AB - In this paper we propose a new ensemble method based on conformal instance transfer. The method combines feature selection and source-instance selection to avoid negative transfer in a model-independent way. It was tested experimentally for different types of classifiers on several benchmark data sets. The experiment results demonstrate that the new method is capable of outperforming significantly standard instance transfer methods.

KW - Instance Transfer

KW - Conformal Prediction

KW - Ensembles

M3 - Conference article in proceeding

VL - 105

T3 - Proceedings of Machine Learning Research

BT - CONFORMAL AND PROBABILISTIC PREDICTION AND APPLICATIONS, VOL 105

PB - JMLR - Journal of Machine Learning Research

T2 - 8th Symposium on Conformal and Probabilistic Prediction and Applications

Y2 - 9 September 2019 through 11 September 2019

ER -

Ensembles based on Conformal Instance Transfer

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this