Bayesian Non-Linear Quantile Regression with Application in Decline Curve Analysis for Petroleum Reservoirs

Total Page:16

File Type:pdf, Size:1020Kb

Bayesian Non-Linear Quantile Regression with Application in Decline Curve Analysis for Petroleum Reservoirs BAYESIAN NON-LINEAR QUANTILE REGRESSION WITH APPLICATION IN DECLINE CURVE ANALYSIS FOR PETROLEUM RESERVOIRS. by YOUJUN LI Submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Mathematics, Applied Mathematics and Statistics CASE WESTERN RESERVE UNIVERSITY May, 2017 CASE WESTERN RESERVE UNIVERSITY SCHOOL OF GRADUATE STUDIES We hereby approve the dissertation of Youjun Li candidate for the degree of Master of Science*. Committee Chair Dr. Anirban Mondal Committee Member Dr. Jenny Brynjarsdottir Committee Member Dr. Wojbor Woyczynski Date of Defense March 31, 2017 *We also certify that written approval has been obtained for any proprietary material contained therein. Contents List of Tables iii List of Figures iv Acknowledgments vi Abstract vii 1 Introduction 1 2 Mean Regression And Quantile Regression Compared (Linear Case) 5 2.1 Mean Regression . .5 2.2 Quantile Regression . .6 2.3 Comparison with Examples . .8 2.3.1 The Household Income Dataset . .8 2.3.2 Simulated Data with Outliers . 12 2.3.3 Econometric Growth Dataset . 14 3 Bayesian Linear Mean and Quantile Regressions 19 3.1 Bayesian Linear Mean Regression . 19 3.2 Bayesian Linear Quantile Regression . 19 3.3 Asymmetric Laplace As Error Distribution . 20 3.4 Bayesian Quantile Regression Demonstrated Using a Real Dataset . 22 i CONTENTS CONTENTS 4 Bayesian Nonlinear Regression with Simulated Oil Reservoir Data 26 4.1 A Brief Description of Decline Curve Analysis . 26 4.2 The Simulated Data with Asymmetric Laplace Error . 27 4.3 Bayesian Nonlinear Mean Regression . 29 4.4 Bayesian Nonlinear Quantile Regression . 29 5 Bayesian Nonlinear Regressions with Real Oil Reservoir Data 38 5.1 Bayesian Model and Sampling from the Posterior . 39 5.1.1 Prior for q0 ............................ 40 5.1.2 Prior for b ............................. 40 5.1.3 Prior for d0 ............................ 40 5.1.4 Prior for σ ............................. 41 5.1.5 The Advantage of \rstan" . 41 5.2 Data Analysis for the First Well . 42 5.3 Data Analysis for the Second Well . 54 6 Discussion 60 7 Future Work 62 8 Bibliography 64 ii List of Tables 3.1 Bayesian Summary of \Prostate Cancer" Dataset . 23 4.1 Summary of Bayesian Median Regression for Simulated Data . 34 5.1 Prior Parameters for q0 .......................... 40 5.2 Prior Parameters for b .......................... 40 5.3 Prior Parameters for d0 .......................... 41 5.4 Prior Parameters for σ .......................... 41 5.5 Median of P-Curvs for Well 1 . 52 5.6 Median of P-Curvs for Well 2 . 55 iii List of Figures 2.1 Graph of the Loss Function ........................7 2.2 Example 1-a Household Income Dataset ..................9 2.3 Example 1-b Household Income Dataset with Mean Regression ...... 10 2.4 Example 1-c Household Income Dataset with Five Quantile Regressions .. 11 2.5 Example 2 Mean and Quantile Regression Against Outliers ........ 13 2.6 Example 3-a GDP and Female Secondary Education ........... 15 2.7 Example 3-b Slopes of Quantile Regressions ................ 16 2.8 Example 3-c All Quantile Regression Lines ................. 17 3.1 Example 4-a Simulated Data with Normal Error .............. 21 3.2 Example 4-b Residual Plots ........................ 22 3.3 Example 5 Traceplot for \Age" ....................... 24 4.1 Example 6 Data Plot ............................ 28 4.2 Example 6 Bayesian Nonlinear Mean Regression Line ........... 30 4.3 Example 6 Traceplots and Histograms ................... 34 4.4 Example 6 Posterior Predictive Curves for Median Regression ....... 37 5.1 Example 7 . 39 5.2 Example 8 . 39 5.3 Example 7-a Traceplots of Mean Regression ................ 43 5.4 Example 7-b Traceplots of Median Regression ............... 44 iv LIST OF FIGURES LIST OF FIGURES 5.5 Example 7-c Traceplots of 10th Quantile Regression ............ 45 5.6 Example 7-d Traceplots of 90th Quantile Regression ............ 46 5.7 Example 7-e Histograms of Mean Regression ................ 47 5.8 Example 7-f Histograms of Median Regression ............... 48 5.9 Example 7-g Histograms of 10th Quantile Regression ........... 49 5.10 Example 7-h Histograms of 90th Quantile Regression ........... 50 5.11 Example 7-i Fitted Curves ......................... 51 5.12 Example 7-j P10 P50 P90 Curves ..................... 53 5.13 Example 7-k Posterior Predictive Curves of Median Regression ...... 55 5.14 Example 8-a Well Two Fitted Curves Compared .............. 56 5.15 Example 8-b Well Two P90 P50 P10 Curves ................ 57 5.16 Example 8-c Well Two Posterior Predictive Curves ............ 59 v Acknowledgments I would like to pay special thankfulness, warmth and appreciation to the persons below who made my study successful and assisted me at every point of the thesis process to cherish my goal: My academic and thesis advisor Dr. Anirban Mondal, for his guidance and exper- tise that made the whole thing possible. My committee member and course instructor Dr. Jenny Brynjarsdottir, for her pro- fessional and kind suggestions that helped me ameliorate the writing of the thesis. My committee member Dr. Wojbor Woyczynski, whose encouragement made me proud of what I have done. My classmate Yuchen Han for reminding me to keep improving. And last but not least, my parents and family, for being supportive no matter what. I am nothing without them. vi Bayesian Non-linear Quantile Regression with Application in Decline Curve Analysis for Petroleum Reservoirs. Abstract by YOUJUN LI In decline curve analysis for hydrocarbon reservoirs, the use of quantile regression instead of the conventional mean regression would be appropriate in the context of oil industry requirement as the fitted quantile regression curves have the correct inter- pretation for the predicted reserves. However, quantiles of a mean regression result have been commonly reported. In this thesis, we consider non-linear quantile regres- sion model where the quantiles of the conditional distribution of the production rate are expressed as some standard non-linear functions of time, under a Bayesian frame- work. The posterior distribution of the regression coefficients and other parameters is intractable mainly due to the non-linearity in the quantile regression function, hence Metropolis Hastings algorithm is used to sample from the posterior. A quantitative assessment of the uncertainty of the decline parameters and the future prediction would be provided for two real datasets. vii 1. Introduction Unlike conventional mean regression, quantile regression was comprehensively stud- ied rather late and not widely used as mean regression. Due to its straightforward intuition, mean regression has dominated the statistical analysis in both business and industry. However, quantile regression managed to handle some difficulties that mean regression failed to overcome. Thanks to the work of Koenker and Bassett Jr (1978), quantile regression be- came an alternative tool to the conventional mean regression in statistical research. In quantile regression models the quantiles of the conditional distribution of the re- sponse variable are expressed as functions of the covariates. The quantile regression is particularly useful when the conditional distribution is heterogeneous and does not have a \standard" shape, such as an asymmetric, fat-tailed, or truncated distribu- tion. Compared to conventional mean regression, quantile regression is more robust to outliers and misspecification of the error distribution. It also provides the interpre- tation of the relationship between different percentiles of the response variable and the predictive variables. On the other hand, Bayesian methodology has been growing rapidly in the digital era. More and more traditional statistical methods have evolved with the integra- tion of Bayesian theory. Kottas and Gelfand (2001) attempted to combine Bayesian approach with quantile regression by considering non-parametric modeling for the error distribution of median regression, a special case of quantile regression, based on 1 Part 1 either P´olya tree or Dirichlet process priors. Yu and Moyeed (2001) further discussed an error distribution modeling from the asymmetric Laplace distribution by proving that the loss function of the quantile regression is distributed as asymmetric Laplace distribution. Although Yu & Moyeed have provided a very thorough study of Bayesian quantile regression, they left the idea of using informative prior for the parameters of the asymmetric Laplace distribution open, such as a prior for the scale parameter σ. In this paper, we will also consider an informative prior for σ with a relatively large variance, so that we can bring the estimate of the error distribution's variance as a part of the posterior inference. Unlike Bayesian quantile regression being applied for linear regression models in the past, application of Bayesian quantile regression for non-linear parametric models have not been well studied in the statistics community. In this thesis, we focus on a particular industrial field, the oil industry, and propose an alternative way to well data analysis other than conventional mean regression. Due to the nature of oil exploitation, the production rate tends to decline over time. Hence the data distribution will not have a \standard" shape, which makes mean regression less adequate. More importantly, the industry regulations require certain percentiles to be reported, indicating that no other approach should be more favorable than quantile regression for this particular problem. As confusing
Recommended publications
  • QUANTILE REGRESSION for CLIMATE DATA Dilhani Marasinghe Clemson University, [email protected]
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Clemson University: TigerPrints Clemson University TigerPrints All Theses Theses 8-2014 QUANTILE REGRESSION FOR CLIMATE DATA Dilhani Marasinghe Clemson University, [email protected] Follow this and additional works at: https://tigerprints.clemson.edu/all_theses Part of the Statistics and Probability Commons Recommended Citation Marasinghe, Dilhani, "QUANTILE REGRESSION FOR CLIMATE DATA" (2014). All Theses. 1909. https://tigerprints.clemson.edu/all_theses/1909 This Thesis is brought to you for free and open access by the Theses at TigerPrints. It has been accepted for inclusion in All Theses by an authorized administrator of TigerPrints. For more information, please contact [email protected]. QUANTILE REGRESSION FOR CLIMATE DATA A Master Thesis Presented to the Graduate School of Clemson University In Partial Fulfillment of the Requirements for the Degree MASTER OF SCIENCE Mathematical Sciences by DILHANI SHALIKA MARASINGHE August 2014 Accepted by: Dr. Collin Gallagher, Committee Chair Dr. Christoper McMahan Dr. Robert Lund Abstract Quantile regression is a developing statistical tool which is used to explain the relationship between response and predictor variables. This thesis describes two examples of climatology using quantile re- gression. Our main goal is to estimate derivatives of a conditional mean and/or conditional quantile function. We introduce a method to handle autocorrelation in the framework of quantile regression and used it with the temperature data. Also we explain some properties of the tornado data which is non-normally distributed. Even though quantile regression provides a more comprehensive view, when talking about residuals with the normality and the constant variance assumption, we would prefer least square regression for our temperature analysis.
    [Show full text]
  • A Toolbox for Nonlinear Regression in R: the Package Nlstools
    JSS Journal of Statistical Software August 2015, Volume 66, Issue 5. http://www.jstatsoft.org/ A Toolbox for Nonlinear Regression in R: The Package nlstools Florent Baty Christian Ritz Sandrine Charles Cantonal Hospital St. Gallen University of Copenhagen University of Lyon Martin Brutsche Jean-Pierre Flandrois Cantonal Hospital St. Gallen University of Lyon Marie-Laure Delignette-Muller University of Lyon Abstract Nonlinear regression models are applied in a broad variety of scientific fields. Various R functions are already dedicated to fitting such models, among which the function nls() has a prominent position. Unlike linear regression fitting of nonlinear models relies on non-trivial assumptions and therefore users are required to carefully ensure and validate the entire modeling. Parameter estimation is carried out using some variant of the least- squares criterion involving an iterative process that ideally leads to the determination of the optimal parameter estimates. Therefore, users need to have a clear understanding of the model and its parameterization in the context of the application and data consid- ered, an a priori idea about plausible values for parameter estimates, knowledge of model diagnostics procedures available for checking crucial assumptions, and, finally, an under- standing of the limitations in the validity of the underlying hypotheses of the fitted model and its implication for the precision of parameter estimates. Current nonlinear regression modules lack dedicated diagnostic functionality. So there is a need to provide users with an extended toolbox of functions enabling a careful evaluation of nonlinear regression fits. To this end, we introduce a unified diagnostic framework with the R package nlstools.
    [Show full text]
  • Gretl User's Guide
    Gretl User’s Guide Gnu Regression, Econometrics and Time-series Allin Cottrell Department of Economics Wake Forest university Riccardo “Jack” Lucchetti Dipartimento di Economia Università Politecnica delle Marche December, 2008 Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.1 or any later version published by the Free Software Foundation (see http://www.gnu.org/licenses/fdl.html). Contents 1 Introduction 1 1.1 Features at a glance ......................................... 1 1.2 Acknowledgements ......................................... 1 1.3 Installing the programs ....................................... 2 I Running the program 4 2 Getting started 5 2.1 Let’s run a regression ........................................ 5 2.2 Estimation output .......................................... 7 2.3 The main window menus ...................................... 8 2.4 Keyboard shortcuts ......................................... 11 2.5 The gretl toolbar ........................................... 11 3 Modes of working 13 3.1 Command scripts ........................................... 13 3.2 Saving script objects ......................................... 15 3.3 The gretl console ........................................... 15 3.4 The Session concept ......................................... 16 4 Data files 19 4.1 Native format ............................................. 19 4.2 Other data file formats ....................................... 19 4.3 Binary databases ..........................................
    [Show full text]
  • An Efficient Nonlinear Regression Approach for Genome-Wide
    An Efficient Nonlinear Regression Approach for Genome-wide Detection of Marginal and Interacting Genetic Variations Seunghak Lee1, Aur´elieLozano2, Prabhanjan Kambadur3, and Eric P. Xing1;? 1School of Computer Science, Carnegie Mellon University, USA 2IBM T. J. Watson Research Center, USA 3Bloomberg L.P., USA [email protected] Abstract. Genome-wide association studies have revealed individual genetic variants associated with phenotypic traits such as disease risk and gene expressions. However, detecting pairwise in- teraction effects of genetic variants on traits still remains a challenge due to a large number of combinations of variants (∼ 1011 SNP pairs in the human genome), and relatively small sample sizes (typically < 104). Despite recent breakthroughs in detecting interaction effects, there are still several open problems, including: (1) how to quickly process a large number of SNP pairs, (2) how to distinguish between true signals and SNPs/SNP pairs merely correlated with true sig- nals, (3) how to detect non-linear associations between SNP pairs and traits given small sam- ple sizes, and (4) how to control false positives? In this paper, we present a unified framework, called SPHINX, which addresses the aforementioned challenges. We first propose a piecewise linear model for interaction detection because it is simple enough to estimate model parameters given small sample sizes but complex enough to capture non-linear interaction effects. Then, based on the piecewise linear model, we introduce randomized group lasso under stability selection, and a screening algorithm to address the statistical and computational challenges mentioned above. In our experiments, we first demonstrate that SPHINX achieves better power than existing methods for interaction detection under false positive control.
    [Show full text]
  • Quantreg: Quantile Regression
    Package ‘quantreg’ June 6, 2021 Title Quantile Regression Description Estimation and inference methods for models of conditional quantiles: Linear and nonlinear parametric and non-parametric (total variation penalized) models for conditional quantiles of a univariate response and several methods for handling censored survival data. Portfolio selection methods based on expected shortfall risk are also now included. See Koenker (2006) <doi:10.1017/CBO9780511754098> and Koenker et al. (2017) <doi:10.1201/9781315120256>. Version 5.86 Maintainer Roger Koenker <[email protected]> Repository CRAN Depends R (>= 2.6), stats, SparseM Imports methods, graphics, Matrix, MatrixModels, conquer Suggests tripack, akima, MASS, survival, rgl, logspline, nor1mix, Formula, zoo, R.rsp License GPL (>= 2) URL https://www.r-project.org NeedsCompilation yes VignetteBuilder R.rsp Author Roger Koenker [cre, aut], Stephen Portnoy [ctb] (Contributions to Censored QR code), Pin Tian Ng [ctb] (Contributions to Sparse QR code), Blaise Melly [ctb] (Contributions to preprocessing code), Achim Zeileis [ctb] (Contributions to dynrq code essentially identical to his dynlm code), Philip Grosjean [ctb] (Contributions to nlrq code), Cleve Moler [ctb] (author of several linpack routines), Yousef Saad [ctb] (author of sparskit2), Victor Chernozhukov [ctb] (contributions to extreme value inference code), Ivan Fernandez-Val [ctb] (contributions to extreme value inference code), Brian D Ripley [trl, ctb] (Initial (2001) R port from S (to my 1 2 R topics documented: everlasting shame -- how could I have been so slow to adopt R!) and for numerous other suggestions and useful advice) Date/Publication 2021-06-06 17:10:02 UTC R topics documented: akj..............................................3 anova.rq . .5 bandwidth.rq .
    [Show full text]
  • Model Selection, Transformations and Variance Estimation in Nonlinear Regression
    Model Selection, Transformations and Variance Estimation in Nonlinear Regression Olaf Bunke1, Bernd Droge1 and J¨org Polzehl2 1 Institut f¨ur Mathematik, Humboldt-Universit¨at zu Berlin PSF 1297, D-10099 Berlin, Germany 2 Konrad-Zuse-Zentrum f¨ur Informationstechnik Heilbronner Str. 10, D-10711 Berlin, Germany Abstract The results of analyzing experimental data using a parametric model may heavily depend on the chosen model. In this paper we propose procedures for the ade- quate selection of nonlinear regression models if the intended use of the model is among the following: 1. prediction of future values of the response variable, 2. estimation of the un- known regression function, 3. calibration or 4. estimation of some parameter with a certain meaning in the corresponding field of application. Moreover, we propose procedures for variance modelling and for selecting an appropriate nonlinear trans- formation of the observations which may lead to an improved accuracy. We show how to assess the accuracy of the parameter estimators by a ”moment oriented bootstrap procedure”. This procedure may also be used for the construction of confidence, prediction and calibration intervals. Programs written in Splus which realize our strategy for nonlinear regression modelling and parameter estimation are described as well. The performance of the selected model is discussed, and the behaviour of the procedures is illustrated by examples. Key words: Nonlinear regression, model selection, bootstrap, cross-validation, variable transformation, variance modelling, calibration, mean squared error for prediction, computing in nonlinear regression. AMS 1991 subject classifications: 62J99, 62J02, 62P10. 1 1 Selection of regression models 1.1 Preliminary discussion In many papers and books it is discussed how to analyse experimental data estimating the parameters in a linear or nonlinear regression model, see e.g.
    [Show full text]
  • Nonlinear Regression, Nonlinear Least Squares, and Nonlinear Mixed Models in R
    Nonlinear Regression, Nonlinear Least Squares, and Nonlinear Mixed Models in R An Appendix to An R Companion to Applied Regression, third edition John Fox & Sanford Weisberg last revision: 2018-06-02 Abstract The nonlinear regression model generalizes the linear regression model by allowing for mean functions like E(yjx) = θ1= f1 + exp[−(θ2 + θ3x)]g, in which the parameters, the θs in this model, enter the mean function nonlinearly. If we assume additive errors, then the parameters in models like this one are often estimated via least squares. In this appendix to Fox and Weisberg (2019) we describe how the nls() function in R can be used to obtain estimates, and briefly discuss some of the major issues with nonlinear least squares estimation. We also describe how to use the nlme() function in the nlme package to fit nonlinear mixed-effects models. Functions in the car package than can be helpful with nonlinear regression are also illustrated. The nonlinear regression model is a generalization of the linear regression model in which the conditional mean of the response variable is not a linear function of the parameters. As a simple example, the data frame USPop in the carData package, which we load along with the car package, has decennial U. S. Census population for the United States (in millions), from 1790 through 2000. The data are shown in Figure 1 (a):1 library("car") Loading required package: carData brief(USPop) 22 x 2 data.frame (17 rows omitted) year population [i] [n] 1 1790 3.9292 2 1800 5.3085 3 1810 7.2399 ..
    [Show full text]
  • 柏際股份有限公司 Bockytech, Inc. 9F-3,No 70, Yanping S
    柏際股份有限公司 BockyTech, Inc. 9F-3,No 70, YanPing S. Rd., Taipei 10042, Taiwan, R.O.C. 10042 台北市延平南路 70 號 9F-3 Http://www.bockytech.com.tw Tel:886-2-23618050 Fax:886-2-23619803 Version 10 LIMDEP Version 10 is an integrated program for estimation and analysis of linear and nonlinear models, with cross section, time series and panel data. LIMDEP has long been a leader in the field of econometric analysis and has provided many recent innovations including cutting edge techniques in panel data analysis, frontier and efficiency estimation and discrete choice modeling. The collection of techniques and procedures for analyzing panel data is without parallel in any other computer program available anywhere. Recognized for years as the standard software for the estimation and manipulation of discrete and limited dependent variable models, LIMDEP10 is now unsurpassed in the breadth and variety of its estimation tools. The main feature of the package is a suite of more than 100 built-in estimators for all forms of the linear regression model, and stochastic frontier, discrete choice and limited dependent variable models, including models for binary, censored, truncated, survival, count, discrete and continuous variables and a variety of sample selection models. No other program offers a wider range of single and multiple equation linear and nonlinear models. LIMDEP is a true state-of-the-art program that is used for teaching and research at thousands of universities, government agencies, research institutes, businesses and industries around the world. LIMDEP is a Complete Econometrics Package LIMDEP takes the form of an econometrics studio.
    [Show full text]
  • Fitting Models to Biological Data Using Linear and Nonlinear Regression
    Fitting Models to Biological Data using Linear and Nonlinear Regression A practical guide to curve fitting Harvey Motulsky & Arthur Christopoulos Copyright 2003 GraphPad Software, Inc. All rights reserved. GraphPad Prism and Prism are registered trademarks of GraphPad Software, Inc. GraphPad is a trademark of GraphPad Software, Inc. Citation: H.J. Motulsky and A Christopoulos, Fitting models to biological data using linear and nonlinear regression. A practical guide to curve fitting. 2003, GraphPad Software Inc., San Diego CA, www.graphpad.com. To contact GraphPad Software, email [email protected] or [email protected]. Contents at a Glance A. Fitting data with nonlinear regression.................................... 13 B. Fitting data with linear regression..........................................47 C. Models ....................................................................................58 D. How nonlinear regression works........................................... 80 E. Confidence intervals of the parameters ..................................97 F. Comparing models................................................................ 134 G. How does a treatment change the curve?..............................160 H. Fitting radioligand and enzyme kinetics data ....................... 187 I. Fitting dose-response curves .................................................256 J. Fitting curves with GraphPad Prism......................................296 3 Contents Preface ........................................................................................................12
    [Show full text]
  • Unconditional Quantile Regressions*
    UNCONDITIONAL QUANTILE REGRESSIONS* Sergio Firpo Nicole M. Fortin Departamento de Economia Department of Economics Catholic University of Rio de Janeiro (PUC-Rio) University of British Columbia R. Marques de S. Vicente, 225/210F #997-1873 East Mall Rio de Janeiro, RJ, Brasil, 22453-900 Vancouver, BC, Canada V6T 1Z1 [email protected] [email protected] Thomas Lemieux Department of Economics University of British Columbia #997-1873 East Mall Vancouver, Canada, BC V6T 1Z1 [email protected] July 2007 ABSTRACT We propose a new regression method to estimate the impact of explanatory variables on quantiles of the unconditional (marginal) distribution of an outcome variable. The proposed method consists of running a regression of the (recentered) influence function (RIF) of the unconditional quantile on the explanatory variables. The influence function is a widely used tool in robust estimation that can easily be computed for each quantile of interest. We show how standard partial effects, as well as policy effects, can be estimated using our regression approach. We propose three different regression estimators based on a standard OLS regression (RIF-OLS), a logit regression (RIF-Logit), and a nonparametric logit regression (RIF-OLS). We also discuss how our approach can be generalized to other distributional statistics besides quantiles. * We are indebted to Joe Altonji, Richard Blundell, David Card, Vinicius Carrasco, Marcelo Fernandes, Chuan Goh, Joel Horowitz, Shakeeb Khan, Roger Koenker, Thierry Magnac, Whitney Newey, Geert Ridder, Jean-Marc Robin, Hal White and seminar participants at Yale University, CAEN-UFC, CEDEPLARUFMG, PUC-Rio, IPEA-RJ and Econometrics in Rio 2006 for useful comments on this and earlier versions of the manuscript.
    [Show full text]
  • Stochastic Dominance Via Quantile Regression
    Abstract We derive a new way to test for stochastic dominance between the return of two assets using a quantile regression formulation. The test statistic is a variant of the one-sided Kolmogorov-Smirnoffstatistic and has a limiting distribution of the standard Brownian bridge. We also illustrate how the test statistic can be extended to test for stochastic dominance among k assets. This is useful when comparing the performance of individual assets in a portfolio against some market index. We show how the test statistic can be modified to test for stochastic dominance up to the α-quantile in situation where the return of one asset does not dominate another over the whole spectrum of the return distribution. Keywords: Quantile regression, stochastic dominance, Brownian bridge, test statis- tic. 1Introduction Stochastic dominance finds applications in many areas. In finance, it is used to assess portfolio diversification, capital structure, bankruptcy risk, and option’s price bound. In welfare economics, it is used to measure income distribution and income inequality In reinsurance coverage, the insured use it to select the best coverage option while the insurers use it to assess whether the options are consistently priced. It is also used to select effective treatment in medicine and selection of the best irrigation system in agriculture. There are two big classes of stochastic dominance tests. The first is based on the inf / sup statistics over the support of the distributions as in McFadden (1989), Klecan, McFadden and McFadden (1991), and Kaur, Rao and Singh (1994). The second class is based on comparison of the distributions over a set of grid points as in Anderson (1996), Dardanoni and Forcina (1998, 1999), and Davidson and Duclos (2000).
    [Show full text]
  • Nonlinear Regression and Nonlinear Least Squares
    Nonlinear Regression and Nonlinear Least Squares Appendix to An R and S-PLUS Companion to Applied Regression John Fox January 2002 1 Nonlinear Regression The normal linear regression model may be written yi = xiβ + εi where xi is a (row) vector of predictors for the ith of n observations, usually with a 1 in the first position representing the regression constant; β is the vector of regression parameters to be estimated; and εi is a random error, assumed to be normally distributed, independently of the errors for other observations, with 2 expectation 0 and constant variance: εi ∼ NID(0,σ ). In the more general normal nonlinear regression model, the function f(·) relating the response to the predictors is not necessarily linear: yi = f(β, xi)+εi As in the linear model, β is a vector of parameters and xi is a vector of predictors (but in the nonlinear 2 regression model, these vectors are not generally of the same dimension), and εi ∼ NID(0,σ ). The likelihood for the nonlinear regression model is n y − f β, x 2 L β,σ2 1 − i=1 i i ( )= n/2 exp 2 (2πσ2) 2σ This likelihood is maximized when the sum of squared residuals n 2 S(β)= yi − f β, xi i=1 is minimized. Differentiating S(β), ∂S β ∂f β, x ( ) − y − f β, x i ∂β = 2 i i ∂β Setting the partial derivatives to 0 produces estimating equations for the regression coefficients. Because these equations are in general nonlinear, they require solution by numerical optimization. As in a linear model, it is usual to estimate the error variance by dividing the residual sum of squares for the model by the number of observations less the number of parameters (in preference to the ML estimator, which divides by n).
    [Show full text]