Temporal and Spatiotemporal Modeling for Multiple Stream Data
DescriptionAccording to Kenneth et al (2004) reviewed findings shows that most of the disease outbreak is detectable by tracking a variety of syndromic data in behavioral patterns, symptoms, and signs. However, most of the existing surveillance algorithms and models are developed by employing standard statistical process control (Burkom et al., 2007), regression, time series, and forecast-based methods (Tsui et al., 2008) and focused on single data stream problems. It is challenging to develop effective methods for timely detection with high identification rates even for standard problems with homogeneous populations and independent distributions of occurrences. Currently, surveillance systems lack the ability to examine the disparate data with diverse conditioned datasets from the hospital database.In this research, we propose the development of effective and efficient surveillance methodologies based on the characteristic of multiple stream infectious disease data, that will significantly improve the detection capability and lower the false alarm rates.
|Effective start/end date||1/10/12 → 7/10/14|