Feature importance: Opening a soil-transmitted helminth machine learning model via SHAP
In the field of landscape epidemiology, the contribution of machine learning (ML) to modeling of epidemiological risk scenarios presents itself as a good alternative. This study aims to break with the ”black box” paradigm that underlies the application of automatic learning techniques by using SHAP...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
KeAi Communications Co., Ltd.
2022-03-01
|
| Series: | Infectious Disease Modelling |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2468042722000045 |
| _version_ | 1797203436401328128 |
|---|---|
| author | Carlos Matias Scavuzzo Juan Manuel Scavuzzo Micaela Natalia Campero Melaku Anegagrie Aranzazu Amor Aramendia Agustín Benito Victoria Periago |
| author_facet | Carlos Matias Scavuzzo Juan Manuel Scavuzzo Micaela Natalia Campero Melaku Anegagrie Aranzazu Amor Aramendia Agustín Benito Victoria Periago |
| author_sort | Carlos Matias Scavuzzo |
| collection | DOAJ |
| description | In the field of landscape epidemiology, the contribution of machine learning (ML) to modeling of epidemiological risk scenarios presents itself as a good alternative. This study aims to break with the ”black box” paradigm that underlies the application of automatic learning techniques by using SHAP to determine the contribution of each variable in ML models applied to geospatial health, using the prevalence of hookworms, intestinal parasites, in Ethiopia, where they are widely distributed; the country bears the third-highest burden of hookworm in Sub-Saharan Africa. XGBoost software was used, a very popular ML model, to fit and analyze the data. The Python SHAP library was used to understand the importance in the trained model, of the variables for predictions. The description of the contribution of these variables on a particular prediction was obtained, using different types of plot methods. The results show that the ML models are superior to the classical statistical models; not only demonstrating similar results but also explaining, by using the SHAP package, the influence and interactions between the variables in the generated models. This analysis provides information to help understand the epidemiological problem presented and provides a tool for similar studies. |
| first_indexed | 2024-04-24T08:19:18Z |
| format | Article |
| id | doaj.art-74a9e4c6389c4a92b40f592540558a8e |
| institution | Directory Open Access Journal |
| issn | 2468-0427 |
| language | English |
| last_indexed | 2024-04-24T08:19:18Z |
| publishDate | 2022-03-01 |
| publisher | KeAi Communications Co., Ltd. |
| record_format | Article |
| series | Infectious Disease Modelling |
| spelling | doaj.art-74a9e4c6389c4a92b40f592540558a8e2024-04-17T02:00:48ZengKeAi Communications Co., Ltd.Infectious Disease Modelling2468-04272022-03-0171262276Feature importance: Opening a soil-transmitted helminth machine learning model via SHAPCarlos Matias Scavuzzo0Juan Manuel Scavuzzo1Micaela Natalia Campero2Melaku Anegagrie3Aranzazu Amor Aramendia4Agustín Benito5Victoria Periago6Instituto de Altos Estudios Espaciales Mario Gulich, Univesidad Nacional de Córdoba-Comisión Nacional de Actividades Espaciales, Argentina; Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Buenos Aires, Argentina; Corresponding author. Instituto de Altos Estudios Espaciales Mario Gulich, Univesidad Nacional de Córdoba-Comisión Nacional de Actividades Espaciales, Spain.Instituto de Altos Estudios Espaciales Mario Gulich, Univesidad Nacional de Córdoba-Comisión Nacional de Actividades Espaciales, ArgentinaInstituto de Altos Estudios Espaciales Mario Gulich, Univesidad Nacional de Córdoba-Comisión Nacional de Actividades Espaciales, Argentina; Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Buenos Aires, ArgentinaFundación Mundo Sano, Madrid, Spain; National Centre for Tropical Medicine, Institute of Health Carlos III, Madrid, SpainFundación Mundo Sano, Madrid, Spain; National Centre for Tropical Medicine, Institute of Health Carlos III, Madrid, SpainNational Centre for Tropical Medicine, Institute of Health Carlos III, Madrid, SpainFundación Mundo Sano, Buenos Aires, Argentina; Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Buenos Aires, ArgentinaIn the field of landscape epidemiology, the contribution of machine learning (ML) to modeling of epidemiological risk scenarios presents itself as a good alternative. This study aims to break with the ”black box” paradigm that underlies the application of automatic learning techniques by using SHAP to determine the contribution of each variable in ML models applied to geospatial health, using the prevalence of hookworms, intestinal parasites, in Ethiopia, where they are widely distributed; the country bears the third-highest burden of hookworm in Sub-Saharan Africa. XGBoost software was used, a very popular ML model, to fit and analyze the data. The Python SHAP library was used to understand the importance in the trained model, of the variables for predictions. The description of the contribution of these variables on a particular prediction was obtained, using different types of plot methods. The results show that the ML models are superior to the classical statistical models; not only demonstrating similar results but also explaining, by using the SHAP package, the influence and interactions between the variables in the generated models. This analysis provides information to help understand the epidemiological problem presented and provides a tool for similar studies.http://www.sciencedirect.com/science/article/pii/S2468042722000045ShapShapleyMachine learningRemote sensingHookwormEthiopia |
| spellingShingle | Carlos Matias Scavuzzo Juan Manuel Scavuzzo Micaela Natalia Campero Melaku Anegagrie Aranzazu Amor Aramendia Agustín Benito Victoria Periago Feature importance: Opening a soil-transmitted helminth machine learning model via SHAP Infectious Disease Modelling Shap Shapley Machine learning Remote sensing Hookworm Ethiopia |
| title | Feature importance: Opening a soil-transmitted helminth machine learning model via SHAP |
| title_full | Feature importance: Opening a soil-transmitted helminth machine learning model via SHAP |
| title_fullStr | Feature importance: Opening a soil-transmitted helminth machine learning model via SHAP |
| title_full_unstemmed | Feature importance: Opening a soil-transmitted helminth machine learning model via SHAP |
| title_short | Feature importance: Opening a soil-transmitted helminth machine learning model via SHAP |
| title_sort | feature importance opening a soil transmitted helminth machine learning model via shap |
| topic | Shap Shapley Machine learning Remote sensing Hookworm Ethiopia |
| url | http://www.sciencedirect.com/science/article/pii/S2468042722000045 |
| work_keys_str_mv | AT carlosmatiasscavuzzo featureimportanceopeningasoiltransmittedhelminthmachinelearningmodelviashap AT juanmanuelscavuzzo featureimportanceopeningasoiltransmittedhelminthmachinelearningmodelviashap AT micaelanataliacampero featureimportanceopeningasoiltransmittedhelminthmachinelearningmodelviashap AT melakuanegagrie featureimportanceopeningasoiltransmittedhelminthmachinelearningmodelviashap AT aranzazuamoraramendia featureimportanceopeningasoiltransmittedhelminthmachinelearningmodelviashap AT agustinbenito featureimportanceopeningasoiltransmittedhelminthmachinelearningmodelviashap AT victoriaperiago featureimportanceopeningasoiltransmittedhelminthmachinelearningmodelviashap |