Ensembles based on conformal instance transfer

Shuang Zhou, Evgueni Smirnov, Gijs Schoenmakers
Proceedings of the Eighth Symposium on Conformal and Probabilistic Prediction and Applications, PMLR 105:23-42, 2019.

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.

Cite this Paper

BibTeX
@InProceedings{pmlr-v105-zhou19a, title = {Ensembles based on conformal instance transfer}, author = {Zhou, Shuang and Smirnov, Evgueni and Schoenmakers, Gijs}, booktitle = {Proceedings of the Eighth Symposium on Conformal and Probabilistic Prediction and Applications}, pages = {23--42}, year = {2019}, editor = {Gammerman, Alex and Vovk, Vladimir and Luo, Zhiyuan and Smirnov, Evgueni}, volume = {105}, series = {Proceedings of Machine Learning Research}, month = {09--11 Sep}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v105/zhou19a/zhou19a.pdf}, url = {https://proceedings.mlr.press/v105/zhou19a.html}, 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.} }
Endnote
%0 Conference Paper %T Ensembles based on conformal instance transfer %A Shuang Zhou %A Evgueni Smirnov %A Gijs Schoenmakers %B Proceedings of the Eighth Symposium on Conformal and Probabilistic Prediction and Applications %C Proceedings of Machine Learning Research %D 2019 %E Alex Gammerman %E Vladimir Vovk %E Zhiyuan Luo %E Evgueni Smirnov %F pmlr-v105-zhou19a %I PMLR %P 23--42 %U https://proceedings.mlr.press/v105/zhou19a.html %V 105 %X 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.
APA
Zhou, S., Smirnov, E. & Schoenmakers, G.. (2019). Ensembles based on conformal instance transfer. Proceedings of the Eighth Symposium on Conformal and Probabilistic Prediction and Applications, in Proceedings of Machine Learning Research 105:23-42 Available from https://proceedings.mlr.press/v105/zhou19a.html.

Related Material