Skip to main content

Data Analytics

Description

Early warning systems must not always rely on geographical proximity for modeling the spread of contagious diseases. Instead, graph structures such as airways or social networks are more adequate in those situations. Nodes, associated to cities, are linked by means of edges, which represent routes between cities. Scan statistics are highly successful for the evaluation of clusters in maps based on geographical proximity. The more flexible neighborhood structure of graphs presents difficulties for the direct usage of scan statistics, due to the highly irregular structures involved. Besides, the traffic intensity between connected nodes plays a significant role which is not usually present in scan statistic based models.

 

Objective 

We describe a model for cluster detection and inference on networks based on the scan statistic. Our aim is to detect as early as possible the appearance of an emerging cluster of syndromes due to a real outbreak (signal) amidst unrelated syndromes (noise).

Submitted by elamb on
Description

Spatial scan finds the most anomalous region that has shown increase in observed counts when compared to the expected baseline. As there can be infinitely many regions to search for, most state-of-the-art algorithms assumes a specific shape of the attack region (circles for Kulldorff and rectangles for Ultra-Fast Spatial Scan Statistics). This assumption might reduce the detection power as real world attacks don't follow standard geometric shapes.

 

Objective

We propose discriminative random field approach for detecting a disease outbreak. Given observed data on a spatial grid, the goal is to label each node as being under attack and non-attack.

Submitted by elamb on
Description

Numerous recent papers have evaluated algorithms for biosurveillance anomaly detection. Common essential problems in the disparate, evolving data environment include trends, day-of-week effects, and other systematic behavior. Public health monitors have expressed the need for modifiable case definitions, requiring monitoring of time series that cannot be modeled in advance. Thus, automated algorithm selection is required. Recent research showed superior predictive performance of the H-W forecasting method compared to regression based predictors applied to syndromic data. This effort discusses extension to a practical monitoring tool, including selection from parametric and initialization settings based on limited data history, selection criteria for routine updating, specification of confidence limits, and validation of the resulting algorithm.

 

Objective

The objective is to develop and evaluate an operational alerting algorithm appropriate for the variety of time series behavior observed in biosurveillance data. The Holt-Winters (H-W) implementation of generalized exponential smoothing, comparable to complex regression models in predictive capability and far easier to specify and adapt, is built into a robust detection method.

Submitted by elamb on
Description

Temporally localized outbreaks occur in the presence of a complex background, greatly complicating both retrospective and real-time detection. Numerous techniques have been proposed for adjusting thresholds to account for this variable background. In this paper, we apply wavelet transforms to detect localized structures in health care time series, using a generalization of many of these viewpoints. A rigorous, nonparametric approach is applied in a general setting to identify coherent outbreaks.

Submitted by elamb on
Description

Multiple or irregularly shaped spatial clusters are often found in disease or syndromic surveillance maps. We develop a novel method to delineate the contours of spatial clusters, especially when there is not a clearly dominating primary cluster, through artificial neural networks. The method may be applied either for maps divided into regions or point data set maps.

Submitted by elamb on
Description

This paper investigates the use of data-adaptive multivariate statistical process control (MSPC) charts for outbreak detection using real-world syndromic data. The widely used EARS [1] methods and other adaptive implementations assume implicitly that nonsta-tionarity and/or the lack of historic data preclude the conventional Phase I/Phase II approach of SPC. This work examines that assumption formally by evaluating and comparing the false alarm rates and sensitivity of adaptive and non-adaptive MSPC charts applied to simulated outbreaks injected into both desea-sonalized and raw data.

Submitted by elamb on
Description

This paper discusses selection of temporal alerting algorithms for syndromic surveillance to achieve reliable detection performance based on statistical properties and the epidemiological context of the input data. We used quantities calculated from brief data history to derive criteria for algorithm selection.

Submitted by elamb on
Description

OBJECTIVE This paper describes a series of data mining techniques used to gather and analyze and disseminate large amounts of data from numerous sources in English as well as Chinese. The objective of the analysis is to attempt to identify locations where the data may indicate a current or future outbreak of the A-H5N1 strain of the flu virus.

Submitted by elamb on