Operational local join count statistics for cluster detection

Anselin, Luc; Li, Xun

doi:10.1007/s10109-019-00299-x

Cited by 68 publications

(62 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Whereas the bivariate local Moran's I accommodate continuous variables measured within areas (i.e. crime rate), bivariate and multivariate joint count statistics have been proposed by Anselin and Xun (2019) to assess the coclustering of two or more binary variables for data that exhibits in-place co-location, or where both variables can have a value of one in a given area, and data that does not exhibit in-place co-location, or where only one of the variables can have a value of one in a given area.…”

Section: Multivariate Cluster Detectionmentioning

confidence: 99%

A Bayesian spatial shared component model for identifying crime-general and crime-specific hotspots

2020

View full text Add to dashboard Cite

The spatial patterning of crime hotspots provides place-based information for the design, allocation, and implementation of crime prevention policies and programmes. However, most spatial hotspot identification methods are univariate, analyse a single crime type, and do not consider if hotspots are shared amongst multiple crime types. This study applies a Bayesian spatial shared component model to identify crime-general and crime-specific hotspots for violent crime and property crime at the small-area scale. The spatial shared component model jointly analyzes both violent crime and property crime and separates the area-specific risks of each crime type into one shared component, which captures the underlying crime-general spatial pattern common to both crime types, and one type-specific component, which captures the crime-specific spatial pattern that diverges from the shared pattern. Crime-general and crime-specific hotspots are classified based on the posterior probability estimates of the shared and type-specific components, respectively. Results show that the crime-general pattern explains approximately 81% of the total variation of violent crime and 70% of the total variation of property crime. Crime-general hotspots are found to be more frequent than crime-specific hotspots, and property crime-specific hotspots are more frequent than violent crimespecific hotspots. Crime-general and crime-specific hotspots are areas that may be targeted with comprehensive initiatives designed for multiple crime types or specialized initiatives designed for a single crime type, respectively.

show abstract

Section: Multivariate Cluster Detectionmentioning

confidence: 99%

A Bayesian spatial shared component model for identifying crime-general and crime-specific hotspots

2020

View full text Add to dashboard Cite

show abstract

“…Most discussions of the law similarly seem to be couched in settings that deal with a single variable. Whereas it is straightforward to generalize the concept of “near” to non‐geographic spaces, such as network spaces (e.g., Sui 2004), it is far more challenging to move the notion of “related” from the univariate to a multivariate context (e.g., witness the difficulties in formulating a multivariate spatial autocorrelation coefficient, as reviewed in Anselin 2019, among others). In this article, we take the perspective that “related” is nothing but a different concept of “near,” but situated in attribute space.…”

Section: Introductionmentioning

confidence: 99%

“…In this article, we take the perspective that “related” is nothing but a different concept of “near,” but situated in attribute space. As in Anselin (2019), we exploit the notion of distance between observations in multivariate attribute space to characterize “related.” Along the same lines, Tobler (2004) also suggested that the principle behind multidimensional scaling “equates similarity with distance,” but we go beyond this context and apply the measure of distance to the attribute space itself as well. In essence, then, Tobler’s law in a multivariate context boils down to assessing the extent to which nearness in geographic space matches nearness in multivariate attribute space, or, equivalently, whether geographic neighbors are also attribute neighbors.…”

Section: Introductionmentioning

confidence: 99%

“…However, the static journal format motivated us to select a small pedagogic example to illustrate the various approaches (allowing for graphical presentation on a journal page). As in Anselin (2019), we use the classic data set on “moral statistics of France” reported in an 1833 essay by André‐Michel Guerry (1833). This data set contains a collection of social indicators (e.g., crime, literacy, suicides, etc.)…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Tobler’s Law in a Multivariate World

Anselin

2020

Geographical Analysis

Self Cite

View full text Add to dashboard Cite

Tobler's first law of geography is widely recognized as reflecting broad empirical realities in geography. Its key concepts of "near" and "related" are intuitive in a univariate setting. However, when moving to the joint consideration of spatial patterns among multiple variables, the combination of attribute similarity and geographical similarity that underlies the concept of spatial autocorrelation is much harder to deal with. This article uses the notion of distance in multiattribute space to explore and visualize the connection between "near" and "related" in a multivariate context. We approach this from a global, local, and regional perspective. We outline a number of ways to combine different visualization techniques and introduce a new local neighbor match test for multivariate local clusters. We illustrate the methods by means of Guerry's classic data set on moral statistics in 1833 France.

show abstract

“…Spatial autocorrelation (analysis of the distribution of MLL incidence across the region) took place using the Moran's I statistic [30], which assesses the degree to which similar values in a dataset correspond with similar locations, therefore identifying the degree of clustering of similar data across the whole dataset. This 'global' analysis is complemented by a Local Indicators of Spatial Association (LISA) analysis, in which the Local Moran's I statistic [31] is calculated in order to decompose the global autocorrelation into local clusters of similar values (e.g. a high value surrounded by several other high values) and outliers (e.g.…”

Section: Data Collection and Analysismentioning

confidence: 99%

A Cross-Sectional Study of Prevalence and Spatial Patterns of Major Limb Loss in the Acholi Sub-Region of Uganda

Atim

Loum

Okello

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: Awareness of residual disabilities amongst people living in countries recovering from prolonged armed conflicts and their socio-economic consequences is increasing. Robust data on the prevalence of such disabilities are important for planning cost effective health services. This study outlines the first community-based prevalence study of Major Limb Loss (MLL) in the Acholi sub region of Northern Uganda. The generic lessons learnt are relevant to many other postconflict societies in Asia and Africa. Methods:A cross sectional survey using random cluster sampling was conducted across 8,000 households in eight districts, of which 7,864 were suitable for analysis. The households were sampled randomly using a high-resolution population model generated using a combination of census data and artificial intelligence. Data were collected using semi-structured questionnaires administered by trained staff, and results were statistically analysed to identify patterns.Results: Data obtained from 7,864 households demonstrated that 47.9% (3,763) of households contained people with disabilities, and 3.0%, (237) of households contained people living with MLL. Of those exhibiting disabilities, the most common types were physical disabilities affecting mobility and other conditions limiting vision or hearing. Our analysis suggests that MLL sufferers are disproportionately male, older and less well educated than the general population. Using the identified prevalence rate of MLL (0.6%) and an estimated population value for the Acholi Sub-Region of 1.9 million, we estimate that there are approximately 11,400 MLL sufferers in the region who require long-term rehabilitation services.Conclusions: This is the first large scale study on the prevalence of MLL following the Ugandan civil war -known for extreme forms of violence, cruelty and mutilation. Our study demonstrates the magnitude of the problem still faced by the affected people, almost 15 years after the end of large scale combat, and the relative paucity of rehabilitation services to meet their needs. Suitable alternative health policy frameworks are required to address these relatively invisible needs.

show abstract

Operational local join count statistics for cluster detection

Cited by 68 publications

References 45 publications

A Bayesian spatial shared component model for identifying crime-general and crime-specific hotspots

A Bayesian spatial shared component model for identifying crime-general and crime-specific hotspots

Tobler’s Law in a Multivariate World

A Cross-Sectional Study of Prevalence and Spatial Patterns of Major Limb Loss in the Acholi Sub-Region of Uganda

Contact Info

Product

Resources

About