Skip to main navigation Skip to search Skip to main content

Outlier Detection with Supervised Learning Method

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Outliers are data points that can affect the quality of data and the results of analysis from data mining. Outlier detection can also be seen as a pre-processing step to find data points that do not properly placed in the data set. Previously outlier detection methods are unsupervised. Today, ODDS provide data set with outlier information as a ground truth for supervised learning. However, outliers are commonly minor in any data set. For supervised learning this will lead to an imbalanced data classification problem. This paper presents the result of popular classification method, k-Nearest neighbor, Centroid Classifier, and Naive Bayes to handle outlier detection task. Even the mentioned methods were not designed to detect outlier, they proved by achieving 81% average sensitivity which is good for further research.

Original languageEnglish
Title of host publicationProceedings of 2019 4th International Conference on Sustainable Information Engineering and Technology, SIET 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages306-309
Number of pages4
ISBN (Electronic)9781728138787
DOIs
Publication statusPublished - Sept 2019
Event4th International Conference on Sustainable Information Engineering and Technology, SIET 2019 - Lombok, Indonesia
Duration: 28 Sept 201930 Sept 2019

Publication series

NameProceedings of 2019 4th International Conference on Sustainable Information Engineering and Technology, SIET 2019

Conference

Conference4th International Conference on Sustainable Information Engineering and Technology, SIET 2019
Country/TerritoryIndonesia
CityLombok
Period28/09/1930/09/19

Keywords

  • classification
  • outlier
  • outlier detection
  • sensitivity
  • supervised learning

Fingerprint

Dive into the research topics of 'Outlier Detection with Supervised Learning Method'. Together they form a unique fingerprint.

Cite this