Incorporation Of Local Information Research Articles

Banks, non-banks, and the incorporation of local information in CMBS loan pricing

Comparing banks to non-bank lenders, we investigate whether the geographical distance between lenders, borrowers, and their properties is reflected in the pricing of US mortgages that were included in US commercial mortgage-backed security (CMBS) pools during the 2000 to 2017 period. The difference in loan spreads when the bank-borrower distance increases from zero to the median of about 700 miles is 10 basis points, and this effect is more pronounced if the loan is collateralized by a riskier property. On the contrary, geographical distance does not seem to have any effect on the loan spread of mortgages granted by non-bank lenders. The difference in loan pricing across originator types (even after controlling for key mortgage and property characteristics) suggests that banks and non-bank lenders have different incentives, lending technologies, and/or different types of borrowers. Our results contribute to the emerging literature on non-bank lender behavior.

Multi-branch deep radial basis function networks for facial emotion recognition

Emotion recognition (ER) from facial images is one of the landmark tasks in affective computing with major developments in the last decade. Initial efforts on ER relied on handcrafted features that were used to characterize facial images and then feed to standard predictive models. Recent methodologies comprise end-to-end trainable deep learning methods that simultaneously learn both, features and predictive model. Perhaps the most successful models are based on convolutional neural networks (CNNs). While these models have excelled at this task, they still fail at capturing local patterns that could emerge in the learning process. We hypothesize these patterns could be captured by variants based on locally weighted learning. Specifically, in this paper we propose a CNN based architecture enhanced with multiple branches formed by radial basis function (RBF) units that aims at exploiting local information at the final stage of the learning process. Intuitively, these RBF units capture local patterns shared by similar instances using an intermediate representation, then the outputs of the RBFs are feed to a softmax layer that exploits this information to improve the predictive performance of the model. This feature could be particularly advantageous in ER as cultural / ethnicity differences may be identified by the local units. We evaluate the proposed method in several ER datasets and show the proposed methodology achieves state-of-the-art in some of them, even when we adopt a pre-trained VGG-Face model as backbone. We show it is the incorporation of local information what makes the proposed model competitive.

A methodology to link national and local information for spatial targeting of ammonia mitigation efforts

The effects of atmospheric nitrogen (N) deposition are evident in terrestrial ecosystems worldwide, with eutrophication and acidification leading to significant changes in species composition. Substantial reductions in N deposition from nitrogen oxides emissions have been achieved in recent decades. By contrast, ammonia (NH3) emissions from agriculture have not decreased substantially and are typically highly spatially variable, making efficient mitigation challenging. One solution is to target NH3 mitigation measures spatially in source landscapes to maximize the benefits for nature conservation. The paper develops an approach to link national scale data and detailed local data to help identify suitable measures for spatial targeting of local sources near designated Special Areas of Conservation (SACs). The methodology combines high-resolution national data on emissions, deposition and source attribution with local data on agricultural management and site conditions.Application of the methodology for the full set of 240 SACs in England found that agriculture contributes ∼45 % of total N deposition. Activities associated with cattle farming represented 54 % of agricultural NH3 emissions within 2 km of the SACs, making them a major contributor to local N deposition, followed by mineral fertiliser application (21 %). Incorporation of local information on agricultural management practices at seven example SACs provided the means to correct outcomes compared with national-scale emission factors. The outcomes show how national scale datasets can provide information on N deposition threats at landscape to national scales, while local-scale information helps to understand the feasibility of mitigation measures, including the impact of detailed spatial targeting on N deposition rates to designated sites.

Confidence Intervals for Nonparametric Curve Estimates

Numerous nonparametric regression methods exist that yield consistent estimators of function curves. Often, one is also interested in constructing confidence intervals for the unknown function. When a function estimate is based on a single global smoothing parameter the resulting confidence intervals may hold their desired confidence level 1 – α on average but because bias in nonparametric estimation is not uniform, they do not hold the desired level uniformly at all design points. Most research in this area has focused on mean squared error properties of the estimator, for example MISE, itself a global measure. In addition, measures like MISE are one step removed from the practical issue of coverage probability. Recent work that focuses on coverage probability has considered only coverage in an average sense, ignoring the important issue of uniformity of coverage across the design space. To deal with this problem, a new estimator is developed which uses a local cross-validation criterion (LCV) to determine a separate smoothing parameter for each design point. The local smoothing parameters are then used to compute the point estimators of the regression curve and the corresponding pointwise confidence intervals. Incorporation of local information through the new method is shown, via Monte Carlo simulation, to yield more uniformly valid pointwise confidence intervals for nonparametric regression curves. Diagnostic plots are developed (Breakout Plots) to visually inspect the degree of uniformity of coverage of the confidence intervals. The approach, here applied to cubic smoothing splines, easily generalizes to many other nonparametric regression estimators. The improved curve estimation is not a solely theoretical improvement such as providing an estimator that has a faster EASE convergence rate but shows its worth empirically by yielding improved coverage probabilities through reliable pointwise confidence intervals.

Recasting alternative agriculture as a management model: The value of adept scaling

Abstract This paper identifies the significance of scale considerations as applied to agricultural resource management. Scaling is the act of defining the spatiotemporal level or levels of interest when attempting problem solving. A goal commonly proposed by the diverse interests working toward an alternative agriculture is sustainability. To date, advocates of sustainability have failed to scale meaningfully both their critique of conventional agriculture and the alternatives they represent. Because sustainability is inherently sensitive to changes in analytic context (i.e., is scale-relative), in the absence of explicit scaling, sustainability is unworkable as a measure of system performance. The importance of scale as applied to the concept of sustainability in hierarchical systems has been noted (Lowrance et al., 1986; Am. J. Altern. Agric., 1: 169–173), however, the implications of this finding have not been confronted in the subsequent literature. Going beyond Lowrance et al. (1986), this paper describes the error that accompanies improperly scaled assessment of agricultural sustainability, and identifies two important considerations that emerge through explicit scaling: (1) the necessity of consistent and timely integration of local information into upper-level decision-making, and (2) the necessity of recognizing the scalar level of the downside of resource allocation trade-offs and addressing them accordingly. The first of these considerations suggests the need for agricultural systems that are resilient to collapse from unpredictable and potentially fundamental disturbance. Through incorporation of local information, it is possible to maintain coherent strategies aimed at upper-level objectives (e.g., biodiversity conservation) while supporting the integrity of lower-level sub-systems (e.g., community economic vitality). The second consideration that arises from explicit scaling indicates the value of developing systems which are buffered against political paralysis caused by negative impacts of resource allocation decisions. Victims, or losers, result from allocation and reallocation functions. These individuals and/or institutions must be addressed. Recent enthusiasm regarding win-win scenarios in many cases is buoyed by scaling error. Explicit recognition of the implications of necessary trade-offs, both positive and negative, promotes the development of mechanisms to support losers. Failure to confront the fact that losers are consistently produced exaggerates the negative impact they have on system performance. We define holistic management as identifying, assimilating, and acting on information in a manner consonant with the contextual spatiotemporal scale of the problem at hand. The fundamental challenge facing agriculture is integration of multiple considerations across levels in the system. Explicitly scaled analysis and action in a holistic setting is proposed as a means by which alternative agriculture can respond to this challenge.