A unified framework for measuring preferences for schools and neighborhoods

This paper develops a framework for estimating household preferences for school and neighborhood attributes in the presence of sorting. It embeds a boundary discontinuity design in a heterogeneous residential choice model, addressing the endogeneity of school and neighborhood characteristics. The model is estimated using restricted-access Census data from a large metropolitan area, yielding a number of new results. First, households are willing to pay less than 1 percent more in house prices - substantially lower than previous estimates - when the average performance of the local school increases by 5 percent. Second, much of the apparent willingness to pay for more educated and wealthier neighbors is explained by the correlation of these sociodemographic measures with unobserved neighborhood quality. Third, neighborhood race is not capitalized directly into housing prices; instead, the negative correlation of neighborhood percent black and housing prices is due entirely to the fact that blacks live in unobservably lower-quality neighborhoods. Finally, there is considerable heterogeneity in preferences for schools and neighbors, with households preferring to self-segregate on the basis of both race and education. © 2007 by The University of Chicago. All rights reserved.

Ecological homogenization of urban USA

A visually apparent but scientifically untested outcome of land-use change is homogenization across urban areas, where neighborhoods in different parts of the country have similar patterns of roads, residential lots, commercial areas, and aquatic features. We hypothesize that this homogenization extends to ecological structure and also to ecosystem functions such as carbon dynamics and microclimate, with continental-scale implications. Further, we suggest that understanding urban homogenization will provide the basis for understanding the impacts of urban land-use change from local to continental scales. Here, we show how multi-scale, multidisciplinary datasets from six metropolitan areas that cover the major climatic regions of the US (Phoenix, AZ; Miami, FL; Baltimore, MD; Boston, MA; Minneapolis-St Paul, MN; and Los Angeles, CA) can be used to determine how household and neighborhood characteristics correlate with land-management practices, land-cover composition, and landscape structure and ecosystem functions at local, regional, and continental scales. © The Ecological Society of America.

Learning from Geometry

Subspaces and manifolds are two powerful models for high dimensional signals. Subspaces model linear correlation and are a good fit to signals generated by physical systems, such as frontal images of human faces and multiple sources impinging at an antenna array. Manifolds model sources that are not linearly correlated, but where signals are determined by a small number of parameters. Examples are images of human faces under different poses or expressions, and handwritten digits with varying styles. However, there will always be some degree of model mismatch between the subspace or manifold model and the true statistics of the source. This dissertation exploits subspace and manifold models as prior information in various signal processing and machine learning tasks.

A near-low-rank Gaussian mixture model measures proximity to a union of linear or affine subspaces. This simple model can effectively capture the signal distribution when each class is near a subspace. This dissertation studies how the pairwise geometry between these subspaces affects classification performance. When model mismatch is vanishingly small, the probability of misclassification is determined by the product of the sines of the principal angles between subspaces. When the model mismatch is more significant, the probability of misclassification is determined by the sum of the squares of the sines of the principal angles. Reliability of classification is derived in terms of the distribution of signal energy across principal vectors. Larger principal angles lead to smaller classification error, motivating a linear transform that optimizes principal angles. This linear transformation, termed TRAIT, also preserves some specific features in each class, being complementary to a recently developed Low Rank Transform (LRT). Moreover, when the model mismatch is more significant, TRAIT shows superior performance compared to LRT.

The manifold model enforces a constraint on the freedom of data variation. Learning features that are robust to data variation is very important, especially when the size of the training set is small. A learning machine with large numbers of parameters, e.g., deep neural network, can well describe a very complicated data distribution. However, it is also more likely to be sensitive to small perturbations of the data, and to suffer from suffer from degraded performance when generalizing to unseen (test) data.

From the perspective of complexity of function classes, such a learning machine has a huge capacity (complexity), which tends to overfit. The manifold model provides us with a way of regularizing the learning machine, so as to reduce the generalization error, therefore mitigate overfiting. Two different overfiting-preventing approaches are proposed, one from the perspective of data variation, the other from capacity/complexity control. In the first approach, the learning machine is encouraged to make decisions that vary smoothly for data points in local neighborhoods on the manifold. In the second approach, a graph adjacency matrix is derived for the manifold, and the learned features are encouraged to be aligned with the principal components of this adjacency matrix. Experimental results on benchmark datasets are demonstrated, showing an obvious advantage of the proposed approaches when the training set is small.

Stochastic optimization makes it possible to track a slowly varying subspace underlying streaming data. By approximating local neighborhoods using affine subspaces, a slowly varying manifold can be efficiently tracked as well, even with corrupted and noisy data. The more the local neighborhoods, the better the approximation, but the higher the computational complexity. A multiscale approximation scheme is proposed, where the local approximating subspaces are organized in a tree structure. Splitting and merging of the tree nodes then allows efficient control of the number of neighbourhoods. Deviation (of each datum) from the learned model is estimated, yielding a series of statistics for anomaly detection. This framework extends the classical {\em changepoint detection} technique, which only works for one dimensional signals. Simulations and experiments highlight the robustness and efficacy of the proposed approach in detecting an abrupt change in an otherwise slowly varying low-dimensional manifold.

Can the Quilombo Model of Collective Land Titling Work in Rio’s Favelas?

Due to their informality, the favelas of Rio de Janeiro are in a precarious position. Though the informal neighborhoods have long served as sites of affordable housing for Rio’s poorest residents, changes within in the city related to public security, mega-events, real estate speculation, and urban revitalization jeopardize their permanence. As one possible solution, this study, conducted for the client Catalytic Communities, investigated collective titling in favelas modeled after quilombos, territories recognized and titled by Brazilian federal law as patrimonies of black cultural traditions.

Food Swamps, Obesity & Health Zoning Restrictions on Fast Food Restaurants

Protecting public health is the most legitimate use of zoning, and yet there is minimal progress in applying it to the obesity problem. Zoning could potentially be used to address both unhealthy and healthy food retailers, but lack of evidence regarding the impact of zoning and public opinion on zoning changes are barriers to implementing zoning restrictions on fast food on a larger scale. My dissertation addresses these gaps in our understanding of health zoning as a policy option for altering built, food environments.

Chapter 1 examines the relationship between food swamps and obesity and whether spatial mapping might be useful in identifying priority geographic areas for zoning interventions. I employ an instrumental variables (IV) strategy to correct for the endogeneity problems associated with food environments, namely that individuals may self-select into certain neighborhoods and may consider food availability in their decision process. I utilize highway exits as a source of exogenous variation .Using secondary data from the USDA Food Environment Atlas, ordinary least squares (OLS) and IV regression models were employed to analyze cross-sectional associations between local food environments and the prevalence of obesity. I find even after controlling for food desert effects, food swamps have a positive, statistically significant effect on adult obesity rates.

Chapter 2 applies theories of message framing and prospect theory to the emerging discussion around health zoning policies targeting food environments and to explore public opinion toward a list of potential zoning restrictions on fast-food restaurants (beyond moratoriums on new establishments). In order to explore causality, I employ an online survey experiment manipulating exposure to vignettes with different message frames about health zoning restrictions with two national samples of adult Americans age 18 and over (N1=2,768 and N2=3,236). The second sample oversamples Black Americans (N=1,000) and individuals with high school as their highest level of education. Respondents were randomly assigned to one of six conditions where they were primed with different message frames about the benefits of zoning restrictions on fast food retailers. Participants were then asked to indicate their support for six zoning policies on a Likert scale. Subjects also answered questions about their food store access, eating behaviors, health status and perceptions of food stores by type.

I find that a message frame about Nutrition and increasing Equity in the food system was particularly effective at increasing support for health zoning policies targeting fast food outlets across policy categories (Conditional, Youth-related, Performance and Incentive) and across racial groups. This finding is consistent with an influential environmental justice scholar’s description of “injustice frames” as effective in mobilizing supporters around environmental issues (Taylor 2000). I extend this rationale to food environment obesity prevention efforts and identify Nutrition combined with Equity frames as an arguably universal campaign strategy for bolstering public support of zoning restrictions on fast food retailers.

Bridging my findings from both Chapters 1 and 2, using food swamps as a spatial metaphor may work to identify priority areas for policy intervention, but only if there is an equitable distribution of resources and mobilization efforts to improve consumer food environments. If the structural forces which ration access to land-use planning persist (arguably including the media as gatekeepers to information and producers of message frames) disparities in obesity are likely to widen.