Analyzing the Performance of Transformers for the Prediction of the Blood Glucose Level Considering Imputation and Smoothing
In this paper we investigate the effect of two preprocessing techniques, data imputation and smoothing, in the prediction of blood glucose level in type 1 diabetes patients, using a novel deep learning model called Transformer. We train three models: XGBoost, a one-dimensional convolutional neural n...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-02-01
|
Series: | Big Data and Cognitive Computing |
Subjects: | |
Online Access: | https://www.mdpi.com/2504-2289/7/1/41 |
_version_ | 1797613504091389952 |
---|---|
author | Edgar Acuna Roxana Aparicio Velcy Palomino |
author_facet | Edgar Acuna Roxana Aparicio Velcy Palomino |
author_sort | Edgar Acuna |
collection | DOAJ |
description | In this paper we investigate the effect of two preprocessing techniques, data imputation and smoothing, in the prediction of blood glucose level in type 1 diabetes patients, using a novel deep learning model called Transformer. We train three models: XGBoost, a one-dimensional convolutional neural network (1D-CNN), and the Transformer model to predict future blood glucose levels for a 30-min horizon using a 60-min time series history in the OhioT1DM dataset. We also compare four methods of handling missing time series data during the model training: hourly mean, linear interpolation, cubic interpolation, and spline interpolation; and two smoothing techniques: Kalman smoothing and smoothing splines. Our experiments show that the Transformer performs better than XGBoost and 1D-CNN when only continuous glucose monitoring (CGM) is used as a predictor, and that it is very competitive against XGBoost when CGM and carbohydrate intake from the meal are used to predict blood glucose level. Overall, our results are more accurate than those appearing in the literature. |
first_indexed | 2024-03-11T06:55:40Z |
format | Article |
id | doaj.art-9afeeb7e75af453cbfb7632773af4f7a |
institution | Directory Open Access Journal |
issn | 2504-2289 |
language | English |
last_indexed | 2024-03-11T06:55:40Z |
publishDate | 2023-02-01 |
publisher | MDPI AG |
record_format | Article |
series | Big Data and Cognitive Computing |
spelling | doaj.art-9afeeb7e75af453cbfb7632773af4f7a2023-11-17T09:37:08ZengMDPI AGBig Data and Cognitive Computing2504-22892023-02-01714110.3390/bdcc7010041Analyzing the Performance of Transformers for the Prediction of the Blood Glucose Level Considering Imputation and SmoothingEdgar Acuna0Roxana Aparicio1Velcy Palomino2Mathematical Sciences Department, University of Puerto Rico at Mayaguez, Mayaguez PR00681, Puerto RicoComputer Science Department, University of Puerto Rico at Bayamon, Bayamon PR00959, Puerto RicoComputing and Information Sciences and Engineering, University of Puerto Rico at Mayaguez, Mayaguez PR00681, Puerto RicoIn this paper we investigate the effect of two preprocessing techniques, data imputation and smoothing, in the prediction of blood glucose level in type 1 diabetes patients, using a novel deep learning model called Transformer. We train three models: XGBoost, a one-dimensional convolutional neural network (1D-CNN), and the Transformer model to predict future blood glucose levels for a 30-min horizon using a 60-min time series history in the OhioT1DM dataset. We also compare four methods of handling missing time series data during the model training: hourly mean, linear interpolation, cubic interpolation, and spline interpolation; and two smoothing techniques: Kalman smoothing and smoothing splines. Our experiments show that the Transformer performs better than XGBoost and 1D-CNN when only continuous glucose monitoring (CGM) is used as a predictor, and that it is very competitive against XGBoost when CGM and carbohydrate intake from the meal are used to predict blood glucose level. Overall, our results are more accurate than those appearing in the literature.https://www.mdpi.com/2504-2289/7/1/41diabetesTransformer1D-CNNXGBoostingglucose predictionimputation |
spellingShingle | Edgar Acuna Roxana Aparicio Velcy Palomino Analyzing the Performance of Transformers for the Prediction of the Blood Glucose Level Considering Imputation and Smoothing Big Data and Cognitive Computing diabetes Transformer 1D-CNN XGBoosting glucose prediction imputation |
title | Analyzing the Performance of Transformers for the Prediction of the Blood Glucose Level Considering Imputation and Smoothing |
title_full | Analyzing the Performance of Transformers for the Prediction of the Blood Glucose Level Considering Imputation and Smoothing |
title_fullStr | Analyzing the Performance of Transformers for the Prediction of the Blood Glucose Level Considering Imputation and Smoothing |
title_full_unstemmed | Analyzing the Performance of Transformers for the Prediction of the Blood Glucose Level Considering Imputation and Smoothing |
title_short | Analyzing the Performance of Transformers for the Prediction of the Blood Glucose Level Considering Imputation and Smoothing |
title_sort | analyzing the performance of transformers for the prediction of the blood glucose level considering imputation and smoothing |
topic | diabetes Transformer 1D-CNN XGBoosting glucose prediction imputation |
url | https://www.mdpi.com/2504-2289/7/1/41 |
work_keys_str_mv | AT edgaracuna analyzingtheperformanceoftransformersforthepredictionofthebloodglucoselevelconsideringimputationandsmoothing AT roxanaaparicio analyzingtheperformanceoftransformersforthepredictionofthebloodglucoselevelconsideringimputationandsmoothing AT velcypalomino analyzingtheperformanceoftransformersforthepredictionofthebloodglucoselevelconsideringimputationandsmoothing |