This paper presents an artificial intelligence-based model, called ANN-2Day model, for forecasting, managing and ultimately eliminating the growing risk of oyster norovirus outbreaks. The ANN-2Day model was developed using Artificial Neural Network (ANN) Toolbox in MATLAB Program and 15-years of epidemiological and environmental data for six independent environmental predictors including water temperature, solar radiation, gage height, salinity, wind, and rainfall. It was found that oyster norovirus outbreaks can be forecasted with two-day lead time using the ANN-2Day model and daily data of the six environmental predictors. Forecasting results of the ANN-2Day model indicated that the model was capable of reproducing 19 years of historical oyster norovirus outbreaks along the Northern Gulf of Mexico coast with the positive predictive value of 76.82%, the negative predictive value of 100.00%, the sensitivity of 100.00%, the specificity of 99.84%, and the overall accuracy of 99.83%, respectively, demonstrating the efficacy of the ANN-2Day model in predicting the risk of norovirus outbreaks to human health. The 2-day lead time enables public health agencies and oyster harvesters to plan for management interventions and thus makes it possible to achieve a paradigm shift of their daily management and operation from primarily reacting to epidemic incidents of norovirus infection after they have occurred to eliminating (or at least reducing) the risk of costly incidents.

本文提出了一种基于人工智能的模型，称为ANN-2Day模型，用于预测，管理并最终消除牡蛎诺如病毒暴发的日益增长的风险。ANN-2Day模型是使用MATLAB程序中的人工神经网络（ANN）工具箱以及15年的流行病学和环境数据开发的，用于六个独立的环境预测因素，包括水温，太阳辐射，计高度，盐度，风和降雨。发现可以使用ANN-2Day模型和六种环境预测因子的每日数据，以两天的交货时间来预测牡蛎诺如病毒的爆发。ANN-2Day模型的预测结果表明，该模型能够重现墨西哥北部湾沿岸19年的牡蛎诺如病毒暴发历史，阳性预测值为76。82％，阴性预测值为100.00％，敏感性为100.00％，特异性为99.84％和整体准确度分别为99.83％，这证明了ANN-2Day模型在预测诺如病毒暴发至人类健康。2天的交货时间使公共卫生机构和牡蛎捕捞者可以计划管理干预措施，从而可以实现其日常管理和操作的范式转变，从最初对诺如病毒感染的流行事件做出反应到消除（或至少降低）发生重大事故的风险。证明了ANN-2Day模型在预测诺如病毒暴发对人类健康的风险方面的功效。2天的交货时间使公共卫生机构和牡蛎捕捞者可以计划管理干预措施，从而可以实现其日常管理和操作的范式转变，从最初对诺如病毒感染的流行事件做出反应到消除（或至少降低）发生重大事故的风险。证明了ANN-2Day模型在预测诺如病毒暴发对人类健康的风险方面的功效。2天的交货时间使公共卫生机构和牡蛎捕捞者可以计划管理干预措施，从而可以实现其日常管理和操作的范式转变，从最初对诺如病毒感染的流行事件做出反应到消除（或至少降低）发生重大事故的风险。