查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – New research on Machine Learning is th e subject of a report. According to news reporting originating from Rocha, Urugu ay, by NewsRx correspondents, research stated, “Faecal contamination is a widesp read environmental and public health problem on recreational beaches around the world. The implementation of predictive models has been recommended by the World Health Organization as a complement to traditional monitoring to assist decisio n-makers and reduce health risks.” Our news editors obtained a quote from the research from Universidad de la Repub lica, “Despite several advances that have been made in the modeling of faecal co liforms, tools and algorithms from machine learning are still scarcely used in t he field and their implementation in nowcast systems is delayed. Here, we perfor m a literature review on modeling strategies to predict faecal contamination in recreational beaches in the last two decades and the implementation of models in nowcast systems to aid management. Models constructed for surface waters of con tinental (lakes, rivers and streams), estuarine and marine coastal ecosystems we re analyzed and compared based on performance metrics for continuous (i.e. regre ssion; R, Root Mean Square Error: RMSE) and categorical (i.e. classification; ac curacy, sensitivity, specificity) responses. We found 67 articles matching the s earch criteria and 40 with information allowing to evaluate and compare predicti ve ability. In early 2000, Multiple Linear Regressions were common, followed by a peak of Artificial Neural Networks (ANNs) from 2010 to 2015, and the rise of M achine learning techniques, such as decision trees (CART and Random Forest) sinc e 2015. ANNs and decision trees presented better accuracy than the remaining mod els. Rainfall and its lags were important predictor variables followed by water temperature. Specificity was much higher than sensitivity in all modeling strate gies, which is typical in data sets where one category (e.g. closed beach) is fa r less common than the normal state (i.e. unbalanced data sets). We registered t he implementation of statistical models in early warning systems in 6 countries, mainly by public beach quality management institutions, followed by NGOs in con junction with universities. We identified critical steps towards improving model construction, evaluation and usage: i) the need to balance the data set previou s to model training, ii) the need to separate data set in training, validation a nd test to perform an honest evaluation of model performance and iii) the transd uction of model outputs to plain language to relevant stakeholders. Integrating into a single framework in situ monitoring, model construction and nowcasting sy stems could help to improve decision making systems to protect users from bathin g in contaminated waters.”
NETL
NSTL