Suchergebnisse

Abstract. Myocardial infarctions (MIs) are a major cause of death worldwide, and both high and low temperatures (i.e. heat and cold) may increase the risk of MI. The relationship between health impacts and climate is complex and influenced by a multitude of climatic, environmental, socio-demographic and behavioural factors. Here, we present a machine learning (ML) approach for predicting MI events based on multiple environmental and demographic variables. We derived data on MI events from the KORA MI registry dataset for Augsburg, Germany, between 1998 and 2015. Multivariable predictors include weather and climate, air pollution (PM10, NO, NO2, SO2 and O3), surrounding vegetation and demographic data. We tested the following ML regression algorithms: decision tree, random forest, multi-layer perceptron, gradient boosting and ridge regression. The models are able to predict the total annual number of MIs reasonably well (adjusted R2=0.62–0.71). Inter-annual variations and long-term trends are captured. Across models the most important predictors are air pollution and daily temperatures. Variables not related to environmental conditions, such as demographics need to be considered as well. This ML approach provides a promising basis to model future MI under changing environmental conditions, as projected by scenarios for climate and other environmental changes.

[EN] In this paper we propose a distributed architecture to provide machine learning practitioners with a set of tools and cloud services that cover the whole machine learning development cycle: ranging from the models creation, training, validation and testing to the models serving as a service, sharing and publication. In such respect, the DEEP-Hybrid-DataCloud framework allows transparent access to existing e-Infrastructures, effectively exploiting distributed resources for the most compute-intensive tasks coming from the machine learning development cycle. Moreover, it provides scientists with a set of Cloud-oriented services to make their models publicly available, by adopting a serverless architecture and a DevOps approach, allowing an easy share, publish and deploy of the developed models. ; This work was supported by the project DEEP-Hybrid-DataCloud ``Designing and Enabling E-infrastructures for intensive Processing in a Hybrid DataCloud'' that has received funding from the European Union's Horizon 2020 Research and Innovation Programme under Grant 777435 ; Lopez Garcia, A.; Marco De Lucas, J.; Antonacci, M.; Zu Castell, W.; David, M.; Hardt, M.; Lloret Iglesias, L. (2020). A Cloud-Based Framework for Machine Learning Workloads and Applications. IEEE Access. 8:18681-18692. https://doi.org/10.1109/ACCESS.2020.2964386 ; S ; 18681 ; 18692 ; 8