A Deep Analysis for Medical Emergency Missing Value Imputation

Md Kabir, Sven Tomforde

2024

Abstract

The prevalence of missing data is a pervasive issue in the medical domain, necessitating the frequent deployment of various imputation techniques. Within the realm of emergency medical care, multiple challenges have been addressed, and solutions have been explored. Notably, the development of an AI assistant for telenotary service (TNA) encounters a significantly higher frequency of missing values compared to other medical applications, with these values missing completely at random. In response to this, we compare several traditional machine learning algorithms with denoising autoencoder and denoising LSTM autoencoder strategies for imputing numerical (continuous) missing values. Our study employs a genuine medical emergency dataset, which is not publicly accessible. This dataset exhibits a significant class imbalance and includes numerous outliers representing rare occurrences. Our findings indicate that the denoising LSTM autoencoder outperforms the conventional approach.

Paper Citation

in Harvard Style

Kabir M. and Tomforde S. (2024). A Deep Analysis for Medical Emergency Missing Value Imputation. In Proceedings of the 16th International Conference on Agents and Artificial Intelligence - Volume 3: ICAART; ISBN 978-989-758-680-4, SciTePress, pages 1229-1236. DOI: 10.5220/0012457300003636

in Bibtex Style

@conference{icaart24,
author={Md Kabir and Sven Tomforde},
title={A Deep Analysis for Medical Emergency Missing Value Imputation},
booktitle={Proceedings of the 16th International Conference on Agents and Artificial Intelligence - Volume 3: ICAART},
year={2024},
pages={1229-1236},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012457300003636},
isbn={978-989-758-680-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 16th International Conference on Agents and Artificial Intelligence - Volume 3: ICAART
TI - A Deep Analysis for Medical Emergency Missing Value Imputation
SN - 978-989-758-680-4
AU - Kabir M.
AU - Tomforde S.
PY - 2024
SP - 1229
EP - 1236
DO - 10.5220/0012457300003636
PB - SciTePress