Conformal feature-selection wrappers for instance transfer

Shuang Zhou; Evgueni Smirnov; Gijs Schoenmakers; Ralf Peeters; Tao Jiang

Conformal feature-selection wrappers for instance transfer

Shuang Zhou, Evgueni Smirnov, Gijs Schoenmakers, Ralf Peeters, Tao Jiang ;

Proceedings of the Seventh Workshop on Conformal and Probabilistic Prediction and Applications, PMLR 91:96-113, 2018.

Abstract

In this paper we propose a new method of conformal feature-selection wrappers for instance transfer (CFSWIT). Given target and source data, the method optimally selects features and source data that are relevant for a classification model. The CFSWIT method is model-independent. It was tested experimentally for several types of classifiers. The experiments show that the CFSWIT method is capable of outperforming standard instance transfer methods.

Related Material

Download PDF

@InProceedings{pmlr-v91-zhou18a, title = {Conformal feature-selection wrappers for instance transfer}, author = {Shuang Zhou and Evgueni Smirnov and Gijs Schoenmakers and Ralf Peeters and Tao Jiang}, booktitle = {Proceedings of the Seventh Workshop on Conformal and Probabilistic Prediction and Applications}, pages = {96--113}, year = {2018}, editor = {Alex Gammerman and Vladimir Vovk and Zhiyuan Luo and Evgueni Smirnov and Ralf Peeters}, volume = {91}, series = {Proceedings of Machine Learning Research}, address = {}, month = {11--13 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v91/zhou18a/zhou18a.pdf}, url = {http://proceedings.mlr.press/v91/zhou18a.html}, abstract = {In this paper we propose a new method of conformal feature-selection wrappers for instance transfer (CFSWIT). Given target and source data, the method optimally selects features and source data that are relevant for a classification model. The CFSWIT method is model-independent. It was tested experimentally for several types of classifiers. The experiments show that the CFSWIT method is capable of outperforming standard instance transfer methods.} }

%0 Conference Paper %T Conformal feature-selection wrappers for instance transfer %A Shuang Zhou %A Evgueni Smirnov %A Gijs Schoenmakers %A Ralf Peeters %A Tao Jiang %B Proceedings of the Seventh Workshop on Conformal and Probabilistic Prediction and Applications %C Proceedings of Machine Learning Research %D 2018 %E Alex Gammerman %E Vladimir Vovk %E Zhiyuan Luo %E Evgueni Smirnov %E Ralf Peeters %F pmlr-v91-zhou18a %I PMLR %J Proceedings of Machine Learning Research %P 96--113 %U http://proceedings.mlr.press %V 91 %W PMLR %X In this paper we propose a new method of conformal feature-selection wrappers for instance transfer (CFSWIT). Given target and source data, the method optimally selects features and source data that are relevant for a classification model. The CFSWIT method is model-independent. It was tested experimentally for several types of classifiers. The experiments show that the CFSWIT method is capable of outperforming standard instance transfer methods.

Zhou, S., Smirnov, E., Schoenmakers, G., Peeters, R. & Jiang, T.. (2018). Conformal feature-selection wrappers for instance transfer. Proceedings of the Seventh Workshop on Conformal and Probabilistic Prediction and Applications, in PMLR 91:96-113