Figure 5 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Euclidean-distance-based LCLQs between motorcycle thefts and entertainment establishments with a bandwidth of: (a) 1 and (b) 25 nearest neighbors [See Figure 2(c) for a bandwidth of 10]; (c) Network-distance-based LCLQ between motorcycle thefts and entertainment establishments with an adaptive bandwidth of 10 nearest neighbor
Source publication
Most existing point-based colocation methods are global measures (e.g., join count statistic, cross K function, and global colocation quotient). Most recently, a local indicator such as the local colocation quotient is proposed to capture the variability of colocation across areas. Our research advances this line of work by developing a simulation-...
Contexts in source publication
Context 1
... facilities) are presented here as motorcycle thefts have the largest sample size. Figures 5A-B show the LCLQ results for bandwidths of 1 and 25 nearest neighbors, respectively. Note that the result for bandwidth of 10 is already shown in Figure 3C. ...
Context 2
... there are significant discrepancies between the two. Figure 5C presents the result of network distance-based LCLQ. Comparing Figure 5C to Figure 3C for the Euclidean distance-based LCLQ, one can spot significant variations in the northwest corner of the study area as marked by the dashed circle, but not in the south (e.g., southeast corner of District A and most part of District B, C and D). ...
Context 3
... 5C presents the result of network distance-based LCLQ. Comparing Figure 5C to Figure 3C for the Euclidean distance-based LCLQ, one can spot significant variations in the northwest corner of the study area as marked by the dashed circle, but not in the south (e.g., southeast corner of District A and most part of District B, C and D). For example, the larger circle in the northwest corner of the study area highlights a cluster of nine motorcycle theft crime incidents that are reported to significantly colocate with entertainment establishments with an average network distance-based LCLQ of 2.15, while insignificant colocation with an average Euclidean distance-based LCLQ of 3.06. ...
Context 4
... there are only a few roads connecting this area with others and the road density around that area is very low; therefore, the distance between an incident and its 10 nearest neighbors would become much longer as measured by network distance than the Euclidean distance, and thus would return results with lower LCLQ values but higher confidence. Similarly, we also find lower LCLQ values by the network distance measure than the Euclidean distance approach in other two areas (in the two marked smaller circles in Figure 5C) where road network is also sparse. This may suggest that the network distance-based LCLQ approach, in general, reports lower but more significant LCLQ values than the Euclidean distance approach, especially in the rural area where road network is sparse and simple. ...
Similar publications
This article analyzes financing alternatives for SMEs in El Salvador (Central America) that operate in the context of high levels of crime. SMEs represent 9.0% of the country's business sector, and together with micro�enterprises they account for 99.6%. Likewise, both participate with 35% of GDP and 67% of employment. The loans from the financial s...
