Ji Meng Loh

Human mobility characterization from cellular network data

Anonymous location data from cellular phone networks sheds light on how people move around on a large scale.

A Tale of One City: Using Cellular Network Data for Urban Planning

Cellular data from call detail records can help urban planners better understand city dynamics. The authors use CDR data to analyze people flow in and out of a suburban city near New York City.

Inequality in obesigenic environments : Fast food density in New York City

The high prevalence of obesity in African American populations may be due to the food environment in residential communities, and the density of fast food restaurants is an important aspect of the restaurant landscape in US cities. This study investigated racial and socioeconomic correlates of fast food density in New York City. We found that predominantly Black areas had higher densities of fast food than predominantly White areas; high-income Black areas had similar exposure as low-income Black areas; and national chains were most dense in commercial areas. The results highlight the importance of policy level interventions to address disparities in food environments as a key goal in obesity prevention efforts.

Route classification using cellular handoff patterns

Understanding utilization of city roads is important for urban planners. In this paper, we show how to use handoff patterns from cellular phone networks to identify which routes people take through a city. Specifically, this paper makes three contributions. First, we show that cellular handoff patterns on a given route are stable across a range of conditions and propose a way to measure stability within and between routes using a variant of Earth Movers Distance. Second, we present two accurate classification algorithms for matching cellular handoff patterns to routes: one requires test drives on the routes while the other uses signal strength data collected by high-resolution scanners. Finally, we present an application of our algorithms for measuring relative volumes of traffic on routes leading into and out of a specific city, and validate our methods using statistics published by a state transportation authority.

A Thinned Block Bootstrap Variance Estimation Procedure for Inhomogeneous Spatial Point Patterns

When modeling inhomogeneous spatial point patterns, it is of interest to fit a parametric model for the first-order intensity function (FOIF) of the process in terms of some measured covariates. Estimates for the regression coefficients, say , can be obtained by maximizing a Poisson maximum likelihood criterion. Little work has been done on the asymptotic distribution of except in some special cases. In this article we show that is asymptotically normal for a general class of mixing processes. To estimate the variance of , we propose a novel thinned block bootstrap procedure that assumes that the point process is second-order reweighted stationary. To apply this procedure, only the FOIF, and not any high-order terms of the process, needs to be estimated. We establish the consistency of the resulting variance estimator, and demonstrate its efficacy through simulations and an application to a real data example.

Accounting for spatial correlation in the scan statistic

The spatial scan statistic is widely used in epidemiology and medical studies as a tool to identify hotspots of diseases. The classical spatial scan statistic assumes the number of disease cases in different locations have independent Poisson distributions, while in practice the data may exhibit overdispersion and spatial correlation. In this work, we examine the behavior of the spatial scan statistic when overdispersion and spatial correlation are present, and propose a modified spatial scan statistic to account for that. Some theoretical results are provided to demonstrate that ignoring the overdispersion and spatial correlation leads to an increased rate of false positives, which is verified through a simulation study. Simulation studies also show that our modified procedure can substantially reduce the rate of false alarms. Two data examples involving brain cancer cases in New Mexico and chickenpox incidence data in France are used to illustrate the practical relevance of the modified procedure.

Statistical distortion: consequences of data cleaning

We introduce the notion of statistical distortion as an essential metric for measuring the effectiveness of data cleaning strategies. We use this metric to propose a widely applicable yet scalable experimental framework for evaluating data cleaning strategies along three dimensions: glitch improvement, statistical distortion and cost-related criteria. Existing metrics focus on glitch improvement and cost, but not on the statistical impact of data cleaning strategies. We illustrate our framework on real world data, with a comprehensive suite of experiments and analyses.

Separate and unequal: The influence of neighborhood and school characteristics on spatial proximity between fast food and schools

OBJECTIVE Social science and health literature have identified residential segregation as a critical factor in exposure to health-related resources, including food environments. Differential spatial patterning of food environments surrounding schools has significant import for youth. We examined whether fast food restaurants clustered around schools in New York City, and whether any observed clustering varied as a function of school type, school racial demographics, and area racial and socioeconomic demographics. METHOD We geocoded fast food locations from 2006 (n=817) and schools from 2004-2005 (n=2096; public and private, elementary and secondary) in the five boroughs of New York City. A point process model (inhomogeneous cross-K function) examined spatial clustering. RESULTS A minimum of 25% of schools had a fast food restaurant within 400 m. High schools had higher fast food clustering than elementary schools. Public elementary and high schools with large proportions of Black students or in block groups with large proportions of Black residents had higher clustering than White counterparts. Finally, public high schools had higher clustering than private counterparts, with 1.25 to 2 times as many restaurants than expected by chance. CONCLUSION The results suggest that the geography of opportunity as it relates to school food environments is unequal in New York City.

A Valid and Fast Spatial Bootstrap for Correlation Functions

In this paper we examine the validity of nonparametric spatial bootstrap as a procedure to quantify errors in estimates of N-point correlation functions. We do this by means of a small simulation study with simple point process models and estimating the two-point correlation functions and their errors. The coverage of confidence intervals obtained using bootstrap is compared with those obtained from assuming Poisson errors. The bootstrap procedure considered here is adapted for use with spatial (i.e., dependent) data. In particular, we describe a marked point bootstrap where, instead of resampling points or blocks of points, we resample marks assigned to the data points. These marks are numerical values that are based on the statistic of interest. We describe how the marks are defined for the two- and three-point correlation functions. By resampling marks, the bootstrap samples retain more of the dependence structure present in the data. Furthermore, this method of bootstrap can be performed much quicker than some other bootstrap methods for spatial data, making it a more practical method with large data sets. We find that with clustered point data sets, confidence intervals obtained using the marked point bootstrap has empirical coverage closer to the nominal level than the confidence intervals obtained using Poisson errors. The bootstrap errors were also found to be closer to the true errors for the clustered point data sets.

Retail Redlining in New York City: Racialized Access to Day-to-Day Retail Resources

Racial residential segregation is associated with health inequalities in the USA, and one of the primary mechanisms is through influencing features of the neighborhood physical environment. To better understand how Black residential segregation might contribute to health risk, we examined retail redlining; the inequitable distribution of retail resources across racially distinct areas. A combination of visual and analytic methods was used to investigate whether predominantly Black census block groups in New York City had poor access to retail stores important for health. After controlling for retail demand, median household income, population density, and subway ridership, percent Black was associated with longer travel distances to various retail industries. Our findings suggest that Black neighborhoods in New York City face retail redlining. Future research is needed to determine how retail redlining may perpetuate health disparities and socioeconomic disadvantage.