Selecting salient objects in real scenes: An oscillatory correlation model

Attention is a critical mechanism for visual scene analysis. By means of attention, it is possible to break down the analysis of a complex scene to the analysis of its parts through a selection process. Empirical studies demonstrate that attentional selection is conducted on visual objects as a whole. We present a neurocomputational model of object-based selection in the framework of oscillatory correlation. By segmenting an input scene and integrating the segments with their conspicuity obtained from a saliency map, the model selects salient objects rather than salient locations. The proposed system is composed of three modules: a saliency map providing saliency values of image locations, image segmentation for breaking the input scene into a set of objects, and object selection which allows one of the objects of the scene to be selected at a time. This object selection system has been applied to real gray-level and color images and the simulation results show the effectiveness of the system. (C) 2010 Elsevier Ltd. All rights reserved.

Empirical analysis of the Lieb-Oxford bound in ions and molecules

Universal properties of the Coulomb interaction energy apply to all many-electron systems. Bounds on the exchange-correlation energy, in particular, are important for the construction of improved density functionals. Here we investigate one such universal property-the Lieb-Oxford lower bound-for ionic and molecular systems. In recent work [J Chem Phys 127, 054106 (2007)], we observed that for atoms and electron liquids this bound may be substantially tightened. Calculations for a few ions and molecules suggested the same tendency, but were not conclusive due to the small number of systems considered. Here we extend that analysis to many different families of ions and molecules, and find that for these, too, the bound can be empirically tightened by a similar margin as for atoms and electron liquids. Tightening the Lieb-Oxford bound will have consequences for the performance of various approximate exchange-correlation functionals. (C) 2008 Wiley Periodicals Inc.

Performance of balanced two-stage empirical predictors of realized cluster latent values from finite populations: A simulation study

Predictors of random effects are usually based on the popular mixed effects (ME) model developed under the assumption that the sample is obtained from a conceptual infinite population; such predictors are employed even when the actual population is finite. Two alternatives that incorporate the finite nature of the population are obtained from the superpopulation model proposed by Scott and Smith (1969. Estimation in multi-stage surveys. J. Amer. Statist. Assoc. 64, 830-840) or from the finite population mixed model recently proposed by Stanek and Singer (2004. Predicting random effects from finite population clustered samples with response error. J. Amer. Statist. Assoc. 99, 1119-1130). Predictors derived under the latter model with the additional assumptions that all variance components are known and that within-cluster variances are equal have smaller mean squared error (MSE) than the competitors based on either the ME or Scott and Smith`s models. As population variances are rarely known, we propose method of moment estimators to obtain empirical predictors and conduct a simulation study to evaluate their performance. The results suggest that the finite population mixed model empirical predictor is more stable than its competitors since, in terms of MSE, it is either the best or the second best and when second best, its performance lies within acceptable limits. When both cluster and unit intra-class correlation coefficients are very high (e.g., 0.95 or more), the performance of the empirical predictors derived under the three models is similar. (c) 2007 Elsevier B.V. All rights reserved.

On the positive correlation between income inequality and unemployment

Several empirical studies in the literature have documented the existence of a positive correlation between income inequalitiy and unemployment. I provide a theoretical framework under which this correlation can be better understood. The analysis is based on a dynamic job search under uncertainty. I start by proving the uniqueness of a stationary distribution of wages in the economy. Drawing upon this distribution, I provide a general expression for the Gini coefficient of income inequality. The expression has the advantage of not requiring a particular specification of the distribution of wage offers. Next, I show how the Gini coefficient varies as a function of the parameters of the model, and how it can be expected to be positively correlated with the rate of unemployment. Two examples are offered. The first, of a technical nature, to show that the convergence of the measures implied by the underlying Markov process can fail in some cases. The second, to provide a quantitative assessment of the model and of the mechanism linking unemployment and inequality.

Non-monotone insurance contracts and their empirical consequences

The goal of this paper is to show the possibility of a non-monotone relation between coverage ans risk which has been considered in the literature of insurance models since the work of Rothschild and Stiglitz (1976). We present an insurance model where the insured agents have heterogeneity in risk aversion and in lenience (a prevention cost parameter). Risk aversion is described by a continuous parameter which is correlated with lenience and for the sake of simplicity, we assume perfect correlation. In the case of positive correlation, the more risk averse agent has higher cosr of prevention leading to a higher demand for coverage. Equivalently, the single crossing property (SCP) is valid and iplies a positive correlation between overage and risk in equilibrium. On the other hand, if the correlation between risk aversion and lenience is negative, not only may the SCP be broken, but also the monotonocity of contracts, i.e., the prediction that high (low) risk averse types choose full (partial) insurance. In both cases riskiness is monotonic in risk aversion, but in the last case there are some coverage levels associated with two different risks (low and high), which implies that the ex-ante (with respect to the risk aversion distribution) correlation between coverage and riskiness may have every sign (even though the ex-post correlation is always positive). Moreover, using another instrument (a proxy for riskiness), we give a testable implication to desentangle single crossing ans non single croosing under an ex-post zero correlation result: the monotonicity of coverage as a function os riskiness. Since by controlling for risk aversion (no asymmetric information), coverage is monotone function of riskiness, this also fives a test for asymmetric information. Finally, we relate this theoretical results to empirical tests in the recent literature, specially the Dionne, Gouruéroux and Vanasse (2001) work. In particular, they found an empirical evidence that seems to be compatible with asymmetric information and non single crossing in our framework. More generally, we build a hidden information model showing how omitted variables (asymmetric information) can bias the sign of the correlation of equilibrium variables conditioning on all observable variables. We show that this may be the case when the omitted variables have a non-monotonic relation with the observable ones. Moreover, because this non-dimensional does not capture this deature. Hence, our main results is to point out the importance of the SPC in testing predictions of the hidden information models.

Non-monotone insurence contracts and their empirical consequences

The goal of t.his paper is to show the possibility of a non-monot.one relation between coverage and risk which has been considered in the literature of insurance models since the work of Rothschild and Stiglitz (1976). We present an insurance model where the insured agents have heterogeneity in risk aversion and in lenience (a prevention cost parameter). Risk aversion is described by a continuou.'l parameter which is correlated with lenience and, for the sake of simplicity, we assume perfect correlation. In the case of positive correlation, the more risk averse agent has higher cost of prevention leading to a higher demand for coverage. Equivalently, the single crossing property (SCP) is valid and implies a positive correlation between coverage and risk in equilibrium. On the other hand, if the correlation between risk aversion and lenience is negative, not only may the sep be broken, but also the monotonicity of contracts, i.e., the prediction that high (Iow) risk averse types choose full (partial) insurance. In both cases riskiness is monotonic in risk aversion, but in the last case t,here are some coverage leveIs associated with two different risks (low and high), which implies that the ex-ante (with respect to the risk aversion distribution) correlation bet,ween coverage and riskiness may have every sign (even though the ex-post correlation is always positive). Moreover, using another instrument (a proxy for riskiness), we give a testable implication to disentangle single crossing and non single crossing under an ex-post zero correlation result: the monotonicity of coverage as a function of riskiness. Since by controlling for risk aversion (no asymmetric informat, ion), coverage is a monotone function of riskiness, this also gives a test for asymmetric information. Finally, we relate this theoretical results to empirica! tests in the recent literature, specially the Dionne, Gouriéroux and Vanasse (2001) work. In particular, they found an empirical evidence that seems to be compatible with asymmetric information and non single crossing in our framework. More generally, we build a hidden information model showing how omitted variabIes (asymmetric information) can bias the sign of the correlation of equilibrium variabIes conditioning on ali observabIe variabIes. We show that this may be t,he case when the omitted variabIes have a non-monotonic reIation with t,he observable ones. Moreover, because this non-monotonic reIat,ion is deepIy reIated with the failure of the SCP in one-dimensional screening problems, the existing lit.erature on asymmetric information does not capture t,his feature. Hence, our main result is to point Out the importance of t,he SCP in testing predictions of the hidden information models.

Interactive Correlation Environment (ICE) - A Statistical Web Tool for Data Collinearity Analysis

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Development of empirical potentials for amorphous silica

Development of empirical potentials for amorphous silica Amorphous silica (SiO2) is of great importance in geoscience and mineralogy as well as a raw material in glass industry. Its structure is characterized as a disordered continuous network of SiO4 tetrahedra. Many efforts have been undertaken to understand the microscopic properties of silica by classical molecular dynamics (MD) simulations. In this method the interatomic interactions are modeled by an effective potential that does not take explicitely into account the electronic degrees of freedom. In this work, we propose a new methodology to parameterize such a potential for silica using ab initio simulations, namely Car-Parrinello (CP) method [Phys. Rev. Lett. 55, 2471 (1985)]. The new potential proposed is compared to the BKS potential [Phys. Rev. Lett. 64, 1955 (1990)] that is considered as the benchmark potential for silica. First, CP simulations have been performed on a liquid silica sample at 3600 K. The structural features so obtained have been compared to the ones predicted by the classical BKS potential. Regarding the bond lengths the BKS tends to underestimate the Si-O bond whereas the Si-Si bond is overestimated. The inter-tetrahedral angular distribution functions are also not well described by the BKS potential. The corresponding mean value of theSiOSi angle is found to be ≃ 147◦, while the CP yields to aSiOSi angle centered around 135◦. Our aim is to fit a classical Born-Mayer/Coulomb pair potential using ab initio calculations. To this end, we use the force-matching method proposed by Ercolessi and Adams [Europhys. Lett. 26, 583 (1994)]. The CP configurations and their corresponding interatomic forces have been considered for a least square fitting procedure. The classical MD simulations with the resulting potential have lead to a structure that is very different from the CP one. Therefore, a different fitting criterion based on the CP partial pair correlation functions was applied. Using this approach the resulting potential shows a better agreement with the CP data than the BKS ones: pair correlation functions, angular distribution functions, structure factors, density of states and pressure/density were improved. At low temperature, the diffusion coefficients appear to be three times higher than those predicted by the BKS model, however showing a similar temperature dependence. Calculations have also been carried out on crystalline samples in order to check the transferability of the potential. The equilibrium geometry as well as the elastic constants of α-quartz at 0 K are well described by our new potential although the crystalline phases have not been considered for the parameterization. We have developed a new potential for silica which represents an improvement over the pair potentials class proposed so far. Furthermore, the fitting methodology that has been developed in this work can be applied to other network forming systems such as germania as well as mixtures of SiO2 with other oxides (e.g. Al2O3, K2O, Na2O).

Complex tracer mobilities in polymer networks revealed by fluorescence correlation spectroscopy

Gels are elastic porous polymer networks that are accompanied by pronounced mechanical properties. Due to their biocompatibility, ‘responsive hydrogels’ (HG) have many biomedical applications ranging from biosensors and drug delivery to tissue engineering. They respond to external stimuli such as temperature and salt by changing their dimensions. Of paramount importance is the ability to engineer penetrability and diffusion of interacting molecules in the crowded HG environment, as this would enable one to optimize a specific functionality. Even though the conditions under which biomedical devices operate are rather complex, a bottom-up approach could reduce the complexity of mutually coupled parameters influencing tracer mobility. The present thesis focuses on the interaction-induced tracer diffusion in polymer solutions and their homologous gels, probed by means of Fluorescence Correlation Spectroscopy (FCS). This is a single-molecule-sensitive technique having the advantage of optimal performance under ultralow tracer concentrations, typically employed in biosensors. Two different types of hydrogels have been investigated, a conventional one with broad polydispersity in the distance between crosslink points and a so-called ‘ideal’, with uniform mesh size distribution. The former is based on a thermoresponsive polymer, exhibiting phase separation in water at temperatures close to the human body temperature. The latter represents an optimal platform to study tracer diffusion. Mobilities of different tracers have been investigated in each network, varying in size, geometry and in terms of tracer-polymer attractive strength, as perturbed by different stimuli. The thesis constitutes a systematic effort towards elucidating the role of the strength and nature of different tracer-polymer interactions, on tracer mobilities; it outlines that interactions can still be very important even in the simplified case of dilute polymer solutions; it also demonstrates that the presence of permanent crosslinks exerts distinct tracer slowdown, depending on the tracer type and the nature of the tracer-polymer interactions, expressed differently by each tracer with regard to the selected stimulus. In aqueous polymer solutions, the tracer slowdown is found to be system-dependent and no universal trend seems to hold, in contrast to predictions from scaling theory for non-interacting nanoparticle mobility and empirical relations concerning the mesh size in polymer solutions. Complex tracer dynamics in polymer networks may be distinctly expressed by FCS, depending on the specific synergy among-at least some of - the following parameters: nature of interactions, external stimuli employed, tracer size and type, crosslink density and swelling ratio.

Near-infrared digital photography to estimate snow correlation length for microwave emission modeling

The study is based on experimental work conducted in alpine snow. We made microwave radiometric and near-infrared reflectance measurements of snow slabs under different experimental conditions. We used an empirical relation to link near-infrared reflectance of snow to the specific surface area (SSA), and converted the SSA into the correlation length. From the measurements of snow radiances at 21 and 35 GHz , we derived the microwave scattering coefficient by inverting two coupled radiative transfer models (the sandwich and six-flux model). The correlation lengths found are in the same range as those determined in the literature using cold laboratory work. The technique shows great potential in the determination of the snow correlation length under field conditions.

Whence a healthy mind: Correlation of physical fitness and academic performance among schoolchildren

Background. Public schools are a key forum in the fight for child health because of the opportunities they present for physical activity and fitness surveillance. However, because schools are evaluated and funded on the basis of standardized academic performance rather than physical activity, empirical research evaluating the connections between fitness and academic performance is needed to justify curriculum allocations to physical activity. ^ Methods. Analyses were based on a convenience sample of 315,092 individually-matched standardized academic (TAKS™) and fitness (FITNESSGRAM®) test records collected by 13 Texas school districts under state mandates. We categorized each fitness result in quintiles by age and gender and used a mixed effects regression model to compare the academic performance of the top and bottom fitness groups for each fitness test and grade level combination. ^ Results. All fitness variables except BMI showed significant, positive associations with academic performance after sociodemographic covariate adjustments, with effect sizes ranging from 0.07 (95% CI: 0.05,0.08) in girls trunklift-TAKS reading to 0.34 (0.32,0.35) in boys cardiovascular-TAKS math. Cardiovascular fitness showed the largest inter-quintile difference in TAKS score (32-75 points), followed by curl-ups. After an additional adjustment for BMI and curl-ups, cardiovascular associations peaked in 8th-9 th grades (maximum inter-quintile difference 142 TAKS points; effect size 0.75 (0.69,0.82) for 8th grade girls math) and showed dose-response characteristics across quintiles (p<0.001 for both genders and outcomes). BMI analysis demonstrated limited, non-linear association with academic performance after adjustment for sociodemographic, cardiovascular fitness and curl-up variables. Low-BMI Hispanic high school boys showed significantly lower TAKS scores than the moderate (but not high) BMI group. High-BMI non-Hispanic white high school girls showed significantly lower scores than the moderate (but not low) BMI group. ^ Conclusions. In this study, fitness was strongly and significantly related to academic performance. Cardiovascular fitness showed a distinct dose-response association with academic performance independent of other sociodemographic and fitness variables. The association peaked in late middle to early high school. The independent association of BMI to academic performance was only found in two sub-groups and was non-linear, with both low and high BMI posing risk relative to moderate BMI but not to each other. In light of our findings, we recommend that policymakers consider PE mandates in middle-high school and require linkage of academic and fitness records to facilitate longitudinal surveillance. School administrators should consider increasing PE time in pursuit of higher academic test scores, and PE practitioners should emphasize cardiovascular fitness over BMI reduction.^

HEALTH MAINTENANCE ORGANIZATION ENVIRONMENT: AN EMPIRICAL STUDY OF THE RELATIONSHIP OF EXTERNAL CONTEXT AND PERFORMANCE

The research project is an extension of a series of administrative science and health care research projects evaluating the influence of external context, organizational strategy, and organizational structure upon organizational success or performance. The research will rely on the assumption that there is not one single best approach to the management of organizations (the contingency theory). As organizational effectiveness is dependent on an appropriate mix of factors, organizations may be equally effective based on differing combinations of factors. The external context of the organization is expected to influence internal organizational strategy and structure and in turn the internal measures affect performance (discriminant theory). The research considers the relationship of external context and organization performance.^ The unit of study for the research will be the health maintenance organization (HMO); an organization the accepts in exchange for a fixed, advance capitation payment, contractual responsibility to assure the delivery of a stated range of health sevices to a voluntary enrolled population. With the current Federal resurgence of interest in the Health Maintenance Organization (HMO) as a major component in the health care system, attention must be directed at maximizing development of HMOs from the limited resources available. Increased skills are needed in both Federal and private evaluation of HMO feasibility in order to prevent resource investment and in projects that will fail while concurrently identifying potentially successful projects that will not be considered using current standards.^ The research considers 192 factors measuring contextual milieu (social, educational, economic, legal, demographic, health and technological factors). Through intercorrelation and principle components data reduction techniques this was reduced to 12 variables. Two measures of HMO performance were identified, they are (1) HMO status (operational or defunct), and (2) a principle components factor score considering eight measures of performance. The relationship between HMO context and performance was analysed using correlation and stepwise multiple regression methods. In each case it has been concluded that the external contextual variables are not predictive of success or failure of study Health Maintenance Organizations. This suggests that performance of an HMO may rely on internal organizational factors. These findings have policy implications as contextual measures are used as a major determinant in HMO feasibility analysis, and as a factor in the allocation of limited Federal funds. ^

Estimation of functional connectivity from electromagnetic signals and the amount of empirical data required

An increasing number of neuroimaging studies are concerned with the identification of interactions or statistical dependencies between brain areas. Dependencies between the activities of different brain regions can be quantified with functional connectivity measures such as the cross-correlation coefficient. An important factor limiting the accuracy of such measures is the amount of empirical data available. For event-related protocols, the amount of data also affects the temporal resolution of the analysis. We use analytical expressions to calculate the amount of empirical data needed to establish whether a certain level of dependency is significant when the time series are autocorrelated, as is the case for biological signals. These analytical results are then contrasted with estimates from simulations based on real data recorded with magnetoencephalography during a resting-state paradigm and during the presentation of visual stimuli. Results indicate that, for broadband signals, 50–100 s of data is required to detect a true underlying cross-correlations coefficient of 0.05. This corresponds to a resolution of a few hundred milliseconds for typical event-related recordings. The required time window increases for narrow band signals as frequency decreases. For instance, approximately 3 times as much data is necessary for signals in the alpha band. Important implications can be derived for the design and interpretation of experiments to characterize weak interactions, which are potentially important for brain processing.

Essays on Firm Heterogeneity with Empirical Applications in Economic History and Agricultural Economics

This thesis uses models of firm-heterogeneity to complete empirical analyses in economic history and agricultural economics. In Chapter 2, a theoretical model of firm heterogeneity is used to derive a statistic that summarizes the welfare gains from the introduction of a new technology. The empirical application considers the use of mechanical steam power in the Canadian manufacturing sector during the late nineteenth century. I exploit exogenous variation in geography to estimate several parameters of the model. My results indicate that the use of steam power resulted in a 15.1 percent increase in firm-level productivity and a 3.0-5.2 percent increase in aggregate welfare. Chapter 3 considers various policy alternatives to price ceiling legislation in the market for production quotas in the dairy farming sector in Quebec. I develop a dynamic model of the demand for quotas with farmers that are heterogeneous in their marginal cost of milk production. The econometric analysis uses farm-level data and estimates a parameter of the theoretical model that is required for the counterfactual experiments. The results indicate that the price of quotas could be reduced to the ceiling price through a 4.16 percent expansion of the aggregate supply of quotas, or through moderate trade liberalization of Canadian dairy products. In Chapter 4, I study the relationship between farm-level productivity and participation in the Commercial Export Milk (CEM) program. I use a difference-in-difference research design with inverse propensity weights to test for causality between participation in the CEM program and total factor productivity (TFP). I find a positive correlation between participation in the CEM program and TFP, however I find no statistically significant evidence that the CEM program affected TFP.