RANK-BASED TEMPO-SPATIAL CLUSTERING: A FRAMEWORK FOR RAPID OUTBREAK DETECTION USING SINGLE OR MULTIPLE DATA STREAMSQue, Jialan (2012) RANK-BASED TEMPO-SPATIAL CLUSTERING: A FRAMEWORK FOR RAPID OUTBREAK DETECTION USING SINGLE OR MULTIPLE DATA STREAMS. Doctoral Dissertation, University of Pittsburgh. (Unpublished)
AbstractIn the recent decades, algorithms for disease outbreak detection have become one of the main interests of public health practitioners to identify and localize an outbreak as early as possible in order to warrant further public health response before a pandemic develops. Today’s increased threat of biological warfare and terrorism provide an even stronger impetus to develop methods for outbreak detection based on symptoms as well as definitive laboratory diagnoses. In this dissertation work, I explore the problems of rapid disease outbreak detection using both spatial and temporal information. I develop a framework of non-parameterized algorithms which search for patterns of disease outbreak in spatial sub-regions of the monitored region within a certain period. Compared to the current existing spatial or tempo-spatial algorithm, the algorithms in this framework provide a methodology for fast searching of either univariate data set or multivariate data set. It first measures which study area is more likely to have an outbreak occurring given the baseline data and currently observed data. Then it applies a greedy searching mechanism to look for clusters with high posterior probabilities given the risk measurement for each unit area as heuristic. I also explore the performance of the proposed algorithms. From the perspective of predictive modeling, I adopt a Gamma-Poisson (GP) model to compute the probability of having an outbreak in each cluster when analyzing univariate data. I build a multinomial generalized Dirichlet (MGD) model to identify outbreak clusters from multivariate data which include the OTC data streams collected by the national retail data monitor (NRDM) and the ED data streams collected by the RODS system. Key contributions of this dissertation include 1) it introduces a rank-based tempo-spatial clustering algorithm, RSC, by utilizing greedy searching and Bayesian GP model for disease outbreak detection with comparable detection timeliness, cluster positive prediction value (PPV) and improved running time; 2) it proposes a multivariate extension of RSC (MRSC) which applies MGD model. The evaluation demonstrated the advantage that MGD model can effectively suppress the false alarms caused by elevated signals that are non-disease relevant and occur in all the monitored data streams. Share
Details
MetricsMonthly Views for the past 3 yearsPlum AnalyticsActions (login required)
|
![[feed]](/images/feed-icon-32x32.png){kind=icon}