Prediction Of Hydrological Models’ Uncertainty By A Committee Of Machine Learning-Models

This study presents an approach to combine uncertainties of the hydrological model outputs predicted from a number of machine learning models. The machine learning based uncertainty prediction approach is very useful for estimation of hydrological models' uncertainty in particular hydro-metrological situation in real-time application [1]. In this approach the hydrological model realizations from Monte Carlo simulations are used to build different machine learning uncertainty models to predict uncertainty (quantiles of pdf) of the a deterministic output from hydrological model . Uncertainty models are trained using antecedent precipitation and streamflows as inputs. The trained models are then employed to predict the model output uncertainty which is specific for the new input data. We used three machine learning models namely artificial neural networks, model tree, locally weighted regression to predict output uncertainties. These three models produce similar verification results, which can be improved by merging their outputs dynamically. We propose an approach to form a committee of the three models to combine their outputs. The approach is applied to estimate uncertainty of streamflows simulation from a conceptual hydrological model in the Brue catchment in UK and the Bagmati catchment in Nepal. The verification results show that merged output is better than an individual model output. [1] D. L. Shrestha, N. Kayastha, and D. P. Solomatine, and R. Price. Encapsulation of parameteric uncertainty statistics by various predictive machine learning models: MLUE method, Journal of Hydroinformatic, in press, 2013.

Uncertainties In Future Projections Of Extreme Rainfall: The Role Of Climate Model, Emission Scenario And Randomness

Climate change has resulted in substantial variations in annual extreme rainfall quantiles in different durations and return periods. Predicting the future changes in extreme rainfall quantiles is essential for various water resources design, assessment, and decision making purposes. Current Predictions of future rainfall extremes, however, exhibit large uncertainties. According to extreme value theory, rainfall extremes are rather random variables, with changing distributions around different return periods; therefore there are uncertainties even under current climate conditions. Regarding future condition, our large-scale knowledge is obtained using global climate models, forced with certain emission scenarios. There are widely known deficiencies with climate models, particularly with respect to precipitation projections. There is also recognition of the limitations of emission scenarios in representing the future global change. Apart from these large-scale uncertainties, the downscaling methods also add uncertainty into estimates of future extreme rainfall when they convert the larger-scale projections into local scale. The aim of this research is to address these uncertainties in future projections of extreme rainfall of different durations and return periods. We plugged 3 emission scenarios with 2 global climate models and used LARS-WG, a well-known weather generator, to stochastically downscale daily climate models’ projections for the city of Saskatoon, Canada, by 2100. The downscaled projections were further disaggregated into hourly resolution using our new stochastic and non-parametric rainfall disaggregator. The extreme rainfall quantiles can be consequently identified for different durations (1-hour, 2-hour, 4-hour, 6-hour, 12-hour, 18-hour and 24-hour) and return periods (2-year, 10-year, 25-year, 50-year, 100-year) using Generalized Extreme Value (GEV) distribution. By providing multiple realizations of future rainfall, we attempt to measure the extent of total predictive uncertainty, which is contributed by climate models, emission scenarios, and downscaling/disaggregation procedures. The results show different proportions of these contributors in different durations and return periods.

Discriminação salarial por cor e gênero revisitada: uma abordagem de decomposição contrafactual utilizando regressões quantílicas

A desigualdade salarial, especialmente a resultante da discriminação contra negros e mulheres no mercado de trabalho, é um componente importante da elevada concentração de renda da economia brasileira. Ao contrário da grande maioria dos trabalhos já desenvolvidos nesta área, este trabalho não adota a hipótese de que os efeitos de atributos determinantes do salário são constantes e idênticos para os indivíduos ao longo da distribuição de renda. São estimadas as estruturas salariais para cada percentil da distribuição salarial para homens brancos, homens negros, mulheres brancas e mulheres negras utilizando a técnica de decomposição contrafactual por regressões quantílicas, proposta por Koenker e Bassett (1978) e desenvolvida por Machado e Mata (2004). Isto proporciona uma compreensão mais detalhada e abrangente dos fatores que determinam a remuneração do trabalho para diferentes níveis de renda e fornece uma medida mais completa do grau de discriminação contra os negros e mulheres no mercado de trabalho ao longo da distribuição salarial. Para os três grupos, a discriminação é crescente em relação à posição na distribuição salarial, indicando a dificuldade de se atingir posições melhor remuneradas no mercado de trabalho por parte de mulheres e negros. A discriminação afeta principalmente as mulheres negras, seguidas das mulheres brancas e dos homens negros. Para os homens negros, a discriminação é baixa entre os mais pobres e cresce nos níveis mais altos da distribuição. As mulheres brancas sofrem ao longo de toda a distribuição com maior efeito entre os 15% mais ricos. As mulheres negras sofrem com a discriminação por cor e gênero, estando assim na pior situação entre os grupos. A remuneração da educação estimada para os quatro grupos indica ganhos crescentes conforme a posição na distribuição salarial ampliando a desigualdade salarial intra-grupo, adicionalmente, observa-se uma desvalorização da educação dos negros de ambos os sexos na determinação salarial e que as mulheres sofrem algum tipo de discriminação no que diz respeito à educação apenas nos níveis salariais mais elevados. Os ganhos salariais obtidos com a equalização da escolaridade e formalização entre os grupos discriminados e os homens brancos indicam que, no caso da educação, homens e mulheres negros teriam ganhos ao longo de toda a distribuição, com ênfase entre os mais ricos. Para a formalização, a população nos decis inferiores da distribuição salarial seria a principal beneficiada.

Do borrowing constraints decrease intergenerational mobility? evidence from Brazil

In this paper, we find evidence that suggests that borrowing constraints may be an important determinant of intergenerational mobility in Brazil. This result contrasts sharply with studies for developed countries, such as Canada and the US, where credit constraints do not seem to play an important role in generating persistence of inequality. Moreover, we find that the social mobility is lower in Brazil in comparison with developed countries. We follow the methodology proposed by Grawe (2001), which uses quantile regression, and obtain two results. First, the degree of intergenerational persistence is greater for the upper quantiles. Second, the degree of intergenerational persistence declines with income at least for the upper quantiles. Both findings are compatible with the presence of borrowing constraints affecting the degree of intergenerational persistence, as predicted by the theory.

Bayesian analysis of extreme events with threshold estimation

The aim of this paper is to analyze extremal events using Generalized Pareto Distributions (GPD), considering explicitly the uncertainty about the threshold. Current practice empirically determines this quantity and proceeds by estimating the GPD parameters based on data beyond it, discarding all the information available be10w the threshold. We introduce a mixture model that combines a parametric form for the center and a GPD for the tail of the distributions and uses all observations for inference about the unknown parameters from both distributions, the threshold inc1uded. Prior distribution for the parameters are indirectly obtained through experts quantiles elicitation. Posterior inference is available through Markov Chain Monte Carlo (MCMC) methods. Simulations are carried out in order to analyze the performance of our proposed mode1 under a wide range of scenarios. Those scenarios approximate realistic situations found in the literature. We also apply the proposed model to a real dataset, Nasdaq 100, an index of the financiai market that presents many extreme events. Important issues such as predictive analysis and model selection are considered along with possible modeling extensions.

Efficient semiparametric estimation of quantile treatment effects

This paper presents calculations of semiparametric efficiency bounds for quantile treatment effects parameters when se1ection to treatment is based on observable characteristics. The paper also presents three estimation procedures forthese parameters, alI ofwhich have two steps: a nonparametric estimation and a computation ofthe difference between the solutions of two distinct minimization problems. Root-N consistency, asymptotic normality, and the achievement ofthe semiparametric efficiency bound is shown for one ofthe three estimators. In the final part ofthe paper, an empirical application to a job training program reveals the importance of heterogeneous treatment effects, showing that for this program the effects are concentrated in the upper quantiles ofthe earnings distribution.

Estudo estatístico sobre eventos de precipitação intensa no nordeste do Brasil

The Northeast of Brazil (NEB) shows high climate variability, ranging from semiarid regions to a rainy regions. According to the latest report of the Intergovernmental Panel on Climate Change, the NEB is highly susceptible to climate change, and also heavy rainfall events (HRE). However, few climatology studies about these episodes were performed, thus the objective main research is to compute the climatology and trend of the episodes number and the daily rainfall rate associated with HRE in the NEB and its climatologically homogeneous sub regions; relate them to the weak rainfall events and normal rainfall events. The daily rainfall data of the hydrometeorological network managed by the Agência Nacional de Águas, from 1972 to 2002. For selection of rainfall events used the technique of quantiles and the trend was identified using the Mann-Kendall test. The sub regions were obtained by cluster analysis, using as similarity measure the Euclidean distance and Ward agglomerative hierarchical method. The results show that the seasonality of the NEB is being intensified, i.e., the dry season is becoming drier and wet season getting wet. The El Niño and La Niña influence more on the amount of events regarding the intensity, but the sub-regions this influence is less noticeable. Using daily data reanalysis ERAInterim fields of anomalies of the composites of meteorological variables were calculated for the coast of the NEB, to characterize the synoptic environment. The Upper-level cyclonic vortex and the South atlantic convergene zone were identified as the main weather systems responsible for training of EPI on the coastland

Evaluating and classifying contaminated areas based on loss functions using annealing simulations

This paper presents a methodology based on geostatistical theory for quantifying the risks associated with heavy-metal contamination in the harbor area of Santana, Amapa State, Northern Brazil. In this area there were activities related to the commercialization of manganese ore from Serra do Navio. Manganese and arsenic concentrations at unsampled sites were estimated by postprocessing results from stochastic annealing simulations; the simulations were used to test different criteria for optimization, including average, median, and quantiles. For classifying areas as contaminated or uncontaminated, estimated quantiles based on functions of asymmetric loss showed better results than did estimates based on symmetric loss, such as the average or the median. The use of specific loss functions in the decision-making process can reduce the costs of remediation and health maintenance. The highest global health costs were observed for manganese. (c) 2008 Elsevier B.V. All rights reserved

Non-stationary frequency analysis of extreme daily rainfall in São Paulo, Brazil

This work is an assessment of frequency of extreme values (EVs) of daily rainfall in the city of São Paulo. Brazil, over the period 1933-2005, based on the peaks-over-threshold (POT) and Generalized Pareto Distribution (GPD) approach. Usually. a GPD model is fitted to a sample of POT Values Selected With a constant threshold. However. in this work we use time-dependent thresholds, composed of relatively large p quantities (for example p of 0.97) of daily rainfall amounts computed from all available data. Samples of POT values were extracted with several Values of p. Four different GPD models (GPD-1, GPD-2, GPD-3. and GDP-4) were fitted to each one of these samples by the maximum likelihood (ML) method. The shape parameter was assumed constant for the four models, but time-varying covariates were incorporated into scale parameter of GPD-2. GPD-3, and GPD-4, describing annual cycle in GPD-2. linear trend in GPD-3, and both annual cycle and linear trend in GPD-4. The GPD-1 with constant scale and shape parameters is the simplest model. For identification of the best model among the four models WC used rescaled Akaike Information Criterion (AIC) with second-order bias correction. This criterion isolates GPD-3 as the best model, i.e. the one with positive linear trend in the scale parameter. The slope of this trend is significant compared to the null hypothesis of no trend, for about 98% confidence level. The non-parametric Mann-Kendall test also showed presence of positive trend in the annual frequency of excess over high thresholds. with p-value being virtually zero. Therefore. there is strong evidence that high quantiles of daily rainfall in the city of São Paulo have been increasing in magnitude and frequency over time. For example. 0.99 quantiles of daily rainfall amount have increased by about 40 mm between 1933 and 2005. Copyright (C) 2008 Royal Meteorological Society

The Poisson-exponential distribution: a Bayesian approach

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Elicitation of multivariate prior distributions: A nonparametric Bayesian approach

In the context of Bayesian statistical analysis, elicitation is the process of formulating a prior density f(.) about one or more uncertain quantities to represent a person's knowledge and beliefs. Several different methods of eliciting prior distributions for one unknown parameter have been proposed. However, there are relatively few methods for specifying a multivariate prior distribution and most are just applicable to specific classes of problems and/or based on restrictive conditions, such as independence of variables. Besides, many of these procedures require the elicitation of variances and correlations, and sometimes elicitation of hyperparameters which are difficult for experts to specify in practice. Garthwaite et al. (2005) discuss the different methods proposed in the literature and the difficulties of eliciting multivariate prior distributions. We describe a flexible method of eliciting multivariate prior distributions applicable to a wide class of practical problems. Our approach does not assume a parametric form for the unknown prior density f(.), instead we use nonparametric Bayesian inference, modelling f(.) by a Gaussian process prior distribution. The expert is then asked to specify certain summaries of his/her distribution, such as the mean, mode, marginal quantiles and a small number of joint probabilities. The analyst receives that information, treating it as a data set D with which to update his/her prior beliefs to obtain the posterior distribution for f(.). Theoretical properties of joint and marginal priors are derived and numerical illustrations to demonstrate our approach are given. (C) 2010 Elsevier B.V. All rights reserved.

The Poisson-exponential lifetime distribution

In this paper we proposed a new two-parameters lifetime distribution with increasing failure rate. The new distribution arises on a latent complementary risk problem base. The properties of the proposed distribution are discussed, including a formal proof of its probability density function and explicit algebraic formulae for its reliability and failure rate functions, quantiles and moments, including the mean and variance. A simple EM-type algorithm for iteratively computing maximum likelihood estimates is presented. The Fisher information matrix is derived analytically in order to obtaining the asymptotic covariance matrix. The methodology is illustrated on a real data set. © 2010 Elsevier B.V. All rights reserved.

Motherhood wage penalties and labour market segmentation: Evidence from Argentina

Includes bibliography.

The performance of transnational corporations: Evidence for the manufacturing industry in Chile

Includes bibliography.

A ocorrência de malária em quatro municípios do estado do Pará, de 1988 a 2005, e sua relação com o desmatamento

O objetivo deste trabalho foi estudar a ocorrência de malária em quatro diferentes regiões representativas do estado do Pará, buscando suas possíveis relações com as taxas de desmatamento. Foi realizado um estudo retrospectivo, com dados secundários, no período de 1988 a 2005, através de casos de malária registrados em quatro municípios do Estado (Anajás, Itaituba, Santana do Araguaia e Viseu), como também das taxas de desmatamento fornecidas pelo PRODES-INPE. Aplicou-se a técnica dos Quantis para se estabelecer cinco categorias ou classes de incidência da malária para cada município, sendo gerado posteriormente um IPA representativo para o Estado. De 1988 até 1994, as curvas de incidência de malária acompanham os números de desmatamento. A partir de 1995, evidenciaram-se anos consecutivos com altos índices de ocorrência da doença logo após os períodos de altas taxas de desmatamento, como registrado nos anos de 1995, 2000 e 2004. Percebeu-se que após a época de intenso desmatamento, os casos de malária variaram entre alto e muito alto no seu padrão de incidência, apontando que o desmatamento pode ser um fator de incremento na frequência e aumento no número de pessoas infectadas no estado do Pará.