Understanding the spatio-temporal characteristics or patterns of the 2019 novel coronavirus (2019-nCoV) epidemic is critical in effectively preventing and controlling this epidemic. However, no research analyzed the spatial dependency and temporal dynamics of 2019-nCoV. Consequently, this research aims to detect the spatio-temporal patterns of the 2019-nCoV epidemic using spatio-temporal analysis methods at the county level in Hubei province. The Mann-Kendall and Pettitt methods were used to identify the temporal trends and abrupt changes in the time series of daily new confirmed cases, respectively. The local Moran's I index was applied to uncover the spatial patterns of the incidence rate, including spatial clusters and outliers. On the basis of the data from January 26 to February 11, 2020, we found that there were 11 areas with different types of temporal patterns of daily new confirmed cases. The pattern characterized by an increasing trend and abrupt change is mainly attributed to the improvement in the ability to diagnose the disease. Spatial clusters with high incidence rates during the period were concentrated in Wuhan Metropolitan Area due to the high intensity of spatial interaction of the population. Therefore, enhancing the ability to diagnose the disease and controlling the movement of the population can be confirmed as effective measures to prevent and control the regional outbreak of the epidemic.Indeed, to effectively prevent and control this epidemic, one of the most critical tasks is to clearly and comprehensively understand the epidemic from different perspectives. Currently, a large body of research has been accomplished, especially in the field of medicine. Mainly, these investigations focus on the genomic characterization of the novel coronavirus [3,4], the clinical characterization of the infected patients [5,6], and the medical diagnosis and treatment of the infected cases [7,8]. The medical research immensely contributes to the recognition of the virus' characteristics and to obtaining the cure. However, no study has discussed the spatio-temporal patterns of the 2019-nCoV epidemic. Mainly, the spatio-temporal patterns describe the epidemical characteristics, including spatio-temporal distribution (regular, clustered, or random), spatio-temporal association, and spatio-temporal evolution [9]. This is beneficial for enhancing the understanding of this epidemic in the spatio-temporal dimension and providing reliable information for decision-making. For example, Liu et al. [10] and Dong et al. [11] analyzed the spatial and temporal characteristics of human infection of avian influenza A(H7N9) in mainland China in 2013. Their results showed that there existed spatially clustered characteristics. Meanwhile, the research by Qiu et al. [12] identified additional risk factors. To understand the dynamic spread of porcine epidemic diarrhea in 2013 in the US, global and local analysis methods uncovered the spatio-temporal patterns of this epidemic [13]. The spatio-temporal distribution and diffusi...

No abstract

Class imbalance ubiquitously exists in real life, which has attracted much interest from various domains. Direct learning from imbalanced dataset may pose unsatisfying results overfocusing on the accuracy of identification and deriving a suboptimal model. Various methodologies have been developed in tackling this problem including sampling, cost-sensitive, and other hybrid ones. However, the samples near the decision boundary which contain more discriminative information should be valued and the skew of the boundary would be corrected by constructing synthetic samples. Inspired by the truth and sense of geometry, we designed a new synthetic minority oversampling technique to incorporate the borderline information. What is more, ensemble model always tends to capture more complicated and robust decision boundary in practice. Taking these factors into considerations, a novel ensemble method, called Bagging of Extrapolation Borderline-SMOTE SVM (BEBS), has been proposed in dealing with imbalanced data learning (IDL) problems. Experiments on open access datasets showed significant superior performance using our model and a persuasive and intuitive explanation behind the method was illustrated. As far as we know, this is the first model combining ensemble of SVMs with borderline information for solving such condition.

No abstract