AutoImpute: An Autonomous Web Tool for Data Imputation Based on Extremely Randomized Trees

Mustafa Alabadla, Fatimah Sidi, Iskandar Ishak, Hamidah Ibrahim, Hazlina Hamdan, Shahril Amir, Appak Nurlankyzy

2023

Abstract

Missing values is one of the main reasons that causes performance degradation, among other things. An inaccurate prediction might result from incorrect imputation of missing variables. A critical step in the study of healthcare information is the imputation of uncertain or missing data. As a result, there has been a significant increase in the development of software tools designed to assist machine learning users in completing their data sets prior to entering them into training algorithms. This study fills the gap by proposing an autonomous imputation application that uses the Extremely Randomised Trees Imputation method to impute mixed-type missing data. The proposed imputation tool provides public users the option to remotely impute their data sets using either of two modes: standard or autonomous. As pointed out in the experimental part, the proposed imputation tool performs better than traditional methods for imputation of missing data on various missing ratios and achieved accurate results for autonomous imputation.

Paper Citation

in Harvard Style

Alabadla M., Sidi F., Ishak I., Ibrahim H., Hamdan H., Amir S. and Nurlankyzy A. (2023). AutoImpute: An Autonomous Web Tool for Data Imputation Based on Extremely Randomized Trees. In Proceedings of the 12th International Conference on Data Science, Technology and Applications - Volume 1: DATA; ISBN 978-989-758-664-4, SciTePress, pages 598-605. DOI: 10.5220/0012144500003541

in Bibtex Style

@conference{data23,
author={Mustafa Alabadla and Fatimah Sidi and Iskandar Ishak and Hamidah Ibrahim and Hazlina Hamdan and Shahril Amir and Appak Nurlankyzy},
title={AutoImpute: An Autonomous Web Tool for Data Imputation Based on Extremely Randomized Trees},
booktitle={Proceedings of the 12th International Conference on Data Science, Technology and Applications - Volume 1: DATA},
year={2023},
pages={598-605},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012144500003541},
isbn={978-989-758-664-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 12th International Conference on Data Science, Technology and Applications - Volume 1: DATA
TI - AutoImpute: An Autonomous Web Tool for Data Imputation Based on Extremely Randomized Trees
SN - 978-989-758-664-4
AU - Alabadla M.
AU - Sidi F.
AU - Ishak I.
AU - Ibrahim H.
AU - Hamdan H.
AU - Amir S.
AU - Nurlankyzy A.
PY - 2023
SP - 598
EP - 605
DO - 10.5220/0012144500003541
PB - SciTePress