On Usefulness of Outlier Elimination in Classification Tasks: Extended Abstract

Dušan Hetlerovič, Luboš Popelı́nský, Pavel Brazdil, Carlos Soares, Fernando Freitas
ECMLPKDD Workshop on Meta-Knowledge Transfer, PMLR 191:78-80, 2022.

Abstract

Although outlier detection/elimination has been studied before, few comprehensive studies exist on when exactly this technique would be useful as preprocessing in classification tasks. Our objective is identify the most useful workflows for a given set of tasks (datasets), and then examine which outlier elimination methods (OEMs) appear in these workflows. The workflows considered in this work are pipelines that include an outlier elimination step followed by a classifier. The OEMs identified this way are considered as useful. Our final aim is to verify what effect this alteration has on generalization performance.

Cite this Paper

BibTeX
@InProceedings{pmlr-v191-hetlerovic22a, title = {On Usefulness of Outlier Elimination in Classification Tasks: Extended Abstract}, author = {Hetlerovi\v{c}, Du\v{s}an and Popel\'{\i}nsk\'{y}, Lubo\v{s} and Brazdil, Pavel and Soares, Carlos and Freitas, Fernando}, booktitle = {ECMLPKDD Workshop on Meta-Knowledge Transfer}, pages = {78--80}, year = {2022}, editor = {Brazdil, Pavel and van Rijn, Jan N. and Gouk, Henry and Mohr, Felix}, volume = {191}, series = {Proceedings of Machine Learning Research}, month = {23 Sep}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v191/hetlerovic22a/hetlerovic22a.pdf}, url = {https://proceedings.mlr.press/v191/hetlerovic22a.html}, abstract = {Although outlier detection/elimination has been studied before, few comprehensive studies exist on when exactly this technique would be useful as preprocessing in classification tasks. Our objective is identify the most useful workflows for a given set of tasks (datasets), and then examine which outlier elimination methods (OEMs) appear in these workflows. The workflows considered in this work are pipelines that include an outlier elimination step followed by a classifier. The OEMs identified this way are considered as useful. Our final aim is to verify what effect this alteration has on generalization performance. } }
Endnote
%0 Conference Paper %T On Usefulness of Outlier Elimination in Classification Tasks: Extended Abstract %A Dušan Hetlerovič %A Luboš Popelı́nský %A Pavel Brazdil %A Carlos Soares %A Fernando Freitas %B ECMLPKDD Workshop on Meta-Knowledge Transfer %C Proceedings of Machine Learning Research %D 2022 %E Pavel Brazdil %E Jan N. van Rijn %E Henry Gouk %E Felix Mohr %F pmlr-v191-hetlerovic22a %I PMLR %P 78--80 %U https://proceedings.mlr.press/v191/hetlerovic22a.html %V 191 %X Although outlier detection/elimination has been studied before, few comprehensive studies exist on when exactly this technique would be useful as preprocessing in classification tasks. Our objective is identify the most useful workflows for a given set of tasks (datasets), and then examine which outlier elimination methods (OEMs) appear in these workflows. The workflows considered in this work are pipelines that include an outlier elimination step followed by a classifier. The OEMs identified this way are considered as useful. Our final aim is to verify what effect this alteration has on generalization performance.
APA
Hetlerovič, D., Popelı́nský, L., Brazdil, P., Soares, C. & Freitas, F.. (2022). On Usefulness of Outlier Elimination in Classification Tasks: Extended Abstract. ECMLPKDD Workshop on Meta-Knowledge Transfer, in Proceedings of Machine Learning Research 191:78-80 Available from https://proceedings.mlr.press/v191/hetlerovic22a.html.

Related Material