Winsorised gini impurity: A resistant to outliers splitting metric for classification tree
Constructing a classification tree is sometimes complicated due to outliers occur in the data. Eliminating the outliers is the simplest option, but some important information will lose. Alternatively, one may make some amendments on the value of outliers, but the amended value is arguable in term of...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Published: |
IP Publishing LLC
2014
|
| Subjects: |
| _version_ | 1825805352704147456 |
|---|---|
| author | Chee, Keong Ch'ng Mahat, Nor Idayu |
| author_facet | Chee, Keong Ch'ng Mahat, Nor Idayu |
| author_sort | Chee, Keong Ch'ng |
| collection | UUM |
| description | Constructing a classification tree is sometimes complicated due to outliers occur in the data. Eliminating the outliers is the simplest option, but some important information will lose. Alternatively, one may make some amendments on the value of outliers, but the amended value is arguable in term of its suitability for classification purposes. We describe a strategy in order to identify and to handle the outliers in the process of constructing a classification tree. A Winsorised approach is suggested in estimating the impurity of the data prior to the splitting of each node of a tree. The proposed estimator provides a splitting value that resistant towards outliers in the data hence influences the performance based on plug in error rate of the tree. We examine the proposed idea on some real data sets represent various sizes of sample. The performance indicates that the proposed strategy is competitive, and sometimes shows better performance than traditional tree. |
| first_indexed | 2024-07-04T06:30:55Z |
| format | Article |
| id | uum-25772 |
| institution | Universiti Utara Malaysia |
| last_indexed | 2024-07-04T06:30:55Z |
| publishDate | 2014 |
| publisher | IP Publishing LLC |
| record_format | eprints |
| spelling | uum-257722019-03-17T03:01:28Z https://repo.uum.edu.my/id/eprint/25772/ Winsorised gini impurity: A resistant to outliers splitting metric for classification tree Chee, Keong Ch'ng Mahat, Nor Idayu QA75 Electronic computers. Computer science Constructing a classification tree is sometimes complicated due to outliers occur in the data. Eliminating the outliers is the simplest option, but some important information will lose. Alternatively, one may make some amendments on the value of outliers, but the amended value is arguable in term of its suitability for classification purposes. We describe a strategy in order to identify and to handle the outliers in the process of constructing a classification tree. A Winsorised approach is suggested in estimating the impurity of the data prior to the splitting of each node of a tree. The proposed estimator provides a splitting value that resistant towards outliers in the data hence influences the performance based on plug in error rate of the tree. We examine the proposed idea on some real data sets represent various sizes of sample. The performance indicates that the proposed strategy is competitive, and sometimes shows better performance than traditional tree. IP Publishing LLC 2014 Article PeerReviewed Chee, Keong Ch'ng and Mahat, Nor Idayu (2014) Winsorised gini impurity: A resistant to outliers splitting metric for classification tree. AIP Conference Proceedings, 1635. pp. 716-723. ISSN 0094-243X http://doi.org/10.1063/1.4903661 doi:10.1063/1.4903661 doi:10.1063/1.4903661 |
| spellingShingle | QA75 Electronic computers. Computer science Chee, Keong Ch'ng Mahat, Nor Idayu Winsorised gini impurity: A resistant to outliers splitting metric for classification tree |
| title | Winsorised gini impurity: A resistant to outliers splitting metric for classification tree |
| title_full | Winsorised gini impurity: A resistant to outliers splitting metric for classification tree |
| title_fullStr | Winsorised gini impurity: A resistant to outliers splitting metric for classification tree |
| title_full_unstemmed | Winsorised gini impurity: A resistant to outliers splitting metric for classification tree |
| title_short | Winsorised gini impurity: A resistant to outliers splitting metric for classification tree |
| title_sort | winsorised gini impurity a resistant to outliers splitting metric for classification tree |
| topic | QA75 Electronic computers. Computer science |
| work_keys_str_mv | AT cheekeongchng winsorisedginiimpurityaresistanttooutlierssplittingmetricforclassificationtree AT mahatnoridayu winsorisedginiimpurityaresistanttooutlierssplittingmetricforclassificationtree |