Missing Data Imputation for Supervised Learning
Missing data imputation can help improve the performance of prediction models in situations where missing data hide useful information. This paper compares methods for imputing missing categorical data for supervised classification tasks. We experiment on two machine learning benchmark datasets with...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2018-04-01
|
| Series: | Applied Artificial Intelligence |
| Online Access: | http://dx.doi.org/10.1080/08839514.2018.1448143 |
| _version_ | 1827817650977243136 |
|---|---|
| author | Jason Poulos Rafael Valle |
| author_facet | Jason Poulos Rafael Valle |
| author_sort | Jason Poulos |
| collection | DOAJ |
| description | Missing data imputation can help improve the performance of prediction models in situations where missing data hide useful information. This paper compares methods for imputing missing categorical data for supervised classification tasks. We experiment on two machine learning benchmark datasets with missing categorical data, comparing classifiers trained on non-imputed (i.e., one-hot encoded) or imputed data with different levels of additional missing-data perturbation. We show imputation methods can increase predictive accuracy in the presence of missing-data perturbation, which can actually improve prediction accuracy by regularizing the classifier. We achieve results comparable to the state-of-the-art on the Adult dataset with missing-data perturbation and $$k$$-nearest-neighbors ($$k$$-NN) imputation. |
| first_indexed | 2024-03-12T00:37:15Z |
| format | Article |
| id | doaj.art-84be66ac51d443ac9c475ee20c821bec |
| institution | Directory Open Access Journal |
| issn | 0883-9514 1087-6545 |
| language | English |
| last_indexed | 2024-03-12T00:37:15Z |
| publishDate | 2018-04-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Applied Artificial Intelligence |
| spelling | doaj.art-84be66ac51d443ac9c475ee20c821bec2023-09-15T09:33:56ZengTaylor & Francis GroupApplied Artificial Intelligence0883-95141087-65452018-04-0132218619610.1080/08839514.2018.14481431448143Missing Data Imputation for Supervised LearningJason Poulos0Rafael Valle1Departments of Political Science and Electrical Engineering and Computer Sciences, University of CaliforniaDepartments of Political Science and Electrical Engineering and Computer Sciences, University of CaliforniaMissing data imputation can help improve the performance of prediction models in situations where missing data hide useful information. This paper compares methods for imputing missing categorical data for supervised classification tasks. We experiment on two machine learning benchmark datasets with missing categorical data, comparing classifiers trained on non-imputed (i.e., one-hot encoded) or imputed data with different levels of additional missing-data perturbation. We show imputation methods can increase predictive accuracy in the presence of missing-data perturbation, which can actually improve prediction accuracy by regularizing the classifier. We achieve results comparable to the state-of-the-art on the Adult dataset with missing-data perturbation and $$k$$-nearest-neighbors ($$k$$-NN) imputation.http://dx.doi.org/10.1080/08839514.2018.1448143 |
| spellingShingle | Jason Poulos Rafael Valle Missing Data Imputation for Supervised Learning Applied Artificial Intelligence |
| title | Missing Data Imputation for Supervised Learning |
| title_full | Missing Data Imputation for Supervised Learning |
| title_fullStr | Missing Data Imputation for Supervised Learning |
| title_full_unstemmed | Missing Data Imputation for Supervised Learning |
| title_short | Missing Data Imputation for Supervised Learning |
| title_sort | missing data imputation for supervised learning |
| url | http://dx.doi.org/10.1080/08839514.2018.1448143 |
| work_keys_str_mv | AT jasonpoulos missingdataimputationforsupervisedlearning AT rafaelvalle missingdataimputationforsupervisedlearning |