Calibrated estimators of the population covariance

Slides:



Advertisements
Similar presentations
Introduction Simple Random Sampling Stratified Random Sampling
Advertisements

Mathematics1 Mathematics 1 Applied Informatics Štefan BEREŽNÝ.
Time (h), x 0123 Distance (m) y Time studying (h), x 0123 grade (%) y
Covariance and Correlation: Estimator/Sample Statistic: Population Parameter: Covariance and correlation measure linear association between two variables,
Nonparametric, Model-Assisted Estimation for a Two-Stage Sampling Design Mark Delorey, F. Jay Breidt, Colorado State University Abstract In aquatic resources,
Bayesian Model Selection and Multi-target Tracking Presenters: Xingqiu Zhao and Nikki Hu Joint work with M. A. Kouritzin, H. Long, J. McCrosky, W. Sun.
1 Def: Let and be random variables of the discrete type with the joint p.m.f. on the space S. (1) is called the mean of (2) is called the variance of (3)
Stochastic Differentiation Lecture 3 Leonidas Sakalauskas Institute of Mathematics and Informatics Vilnius, Lithuania EURO Working Group on Continuous.
Point estimation, interval estimation
Maximum likelihood Conditional distribution and likelihood Maximum likelihood estimations Information in the data and likelihood Observed and Fisher’s.
1 Introduction to Biostatistics (PUBHLTH 540) Multiple Random Variables.
REGRESSION What is Regression? What is the Regression Equation? What is the Least-Squares Solution? How is Regression Based on Correlation? What are the.
Chapter 4 Multiple Regression.
Econ 140 Lecture 71 Classical Regression Lecture 7.
2. Point and interval estimation Introduction Properties of estimators Finite sample size Asymptotic properties Construction methods Method of moments.
1 Sampling Models for the Population Mean Ed Stanek UMASS Amherst.
Continuous Random Variables and Probability Distributions
Basic Mathematics for Portfolio Management. Statistics Variables x, y, z Constants a, b Observations {x n, y n |n=1,…N} Mean.
1 In a second variation, we shall consider the model shown above. x is the rate of growth of productivity, assumed to be exogenous. w is now hypothesized.
Copyright © Cengage Learning. All rights reserved. 3.5 Hypergeometric and Negative Binomial Distributions.
Copyright © 2010, 2007, 2004 Pearson Education, Inc. All Rights Reserved Section 10-1 Review and Preview.
Geo479/579: Geostatistics Ch12. Ordinary Kriging (1)
© 2001 Prentice-Hall, Inc. Statistics for Business and Economics Simple Linear Regression Chapter 10.
CSDA Conference, Limassol, 2005 University of Medicine and Pharmacy “Gr. T. Popa” Iasi Department of Mathematics and Informatics Gabriel Dimitriu University.
Chapter 4 DeGroot & Schervish. Variance Although the mean of a distribution is a useful summary, it does not convey very much information about the distribution.
The Mean of a Discrete RV The mean of a RV is the average value the RV takes over the long-run. –The mean of a RV is analogous to the mean of a large population.
SUPA Advanced Data Analysis Course, Jan 6th – 7th 2009 Advanced Data Analysis for the Physical Sciences Dr Martin Hendry Dept of Physics and Astronomy.
Miscellaneous examples 1. Given that a = log 10 2 and b = log 10 5, find expressions in terms of a and b for: (i) log (ii) log (iii) log 10.
Simple Linear Regression. The term linear regression implies that  Y|x is linearly related to x by the population regression equation  Y|x =  +  x.
Copyright 2010, The World Bank Group. All Rights Reserved. Part 2 Sample Design Produced in Collaboration between World Bank Institute and the Development.
Probability and Statistics
Clustering and Testing in High- Dimensional Data M. Radavičius, G. Jakimauskas, J. Sušinskas (Institute of Mathematics and Informatics, Vilnius, Lithuania)
Eurostat Statistical matching when samples are drawn according to complex survey designs Training Course «Statistical Matching» Rome, 6-8 November 2013.
1 Sample Geometry and Random Sampling Shyh-Kang Jeng Department of Electrical Engineering/ Graduate Institute of Communication/ Graduate Institute of Networking.
Eurostat Weighting and Estimation. Presented by Loredana Di Consiglio Istituto Nazionale di Statistica, ISTAT.
Monte-Carlo method for Two-Stage SLP Lecture 5 Leonidas Sakalauskas Institute of Mathematics and Informatics Vilnius, Lithuania EURO Working Group on Continuous.
Lecture 22 Numerical Analysis. Chapter 5 Interpolation.
Estimation Method of Moments (MM) Methods of Moment estimation is a general method where equations for estimating parameters are found by equating population.
Estimation of covariance matrix under informative sampling Julia Aru University of Tartu and Statistics Estonia Tartu, June 25-29, 2007.
Systems of Linear Equations in Two Variables. 1. Determine whether the given ordered pair is a solution of the system.
Section 1.6 Fitting Linear Functions to Data. Consider the set of points {(3,1), (4,3), (6,6), (8,12)} Plot these points on a graph –This is called a.
1 1 Slide The Simple Linear Regression Model n Simple Linear Regression Model y =  0 +  1 x +  n Simple Linear Regression Equation E( y ) =  0 + 
Econometrics III Evgeniya Anatolievna Kolomak, Professor.
Small area estimation combining information from several sources Jae-Kwang Kim, Iowa State University Seo-Young Kim, Statistical Research Institute July.
Multi-Dimensional Credibility Excess Work Comp Application.
Calculus continued The reverse of differentiation The reverse process of differentiation is called Integration.
Section Copyright © 2015, 2011, 2008 Pearson Education, Inc. Lecture Slides Essentials of Statistics 5 th Edition and the Triola Statistics Series.
Virtual University of Pakistan
12. Principles of Parameter Estimation
CH 5: Multivariate Methods
Chapter 5 STATISTICS (PART 4).
SIMPLE LINEAR REGRESSION MODEL
CORRELATION ANALYSIS.
Two-Phase Sampling (Double Sampling)
Decomposition Methods
Linear Systems.
Lecture Slides Elementary Statistics Twelfth Edition
T test.
Chapter 8: Weighting adjustment
Introduction to Statistical Methods for Measuring “Omics” and Field Data PCA, PcoA, distance measure, AMOVA.
Lecture Slides Elementary Statistics Twelfth Edition
Istat - Structural Business Statistics
Multivariate Methods Berlin Chen
The European Statistical Training Programme (ESTP)
I can determine the different sampling techniques used in real life.
12. Principles of Parameter Estimation
Propagation of Error Berlin Chen
Propagation of Error Berlin Chen
Canonical Correlation Analysis
X ⦁ X = 64 ±8 ±14 X ⦁ X ⦁ X =
Presentation transcript:

Calibrated estimators of the population covariance Aleksandras Plikusas and Dalius Pumputis, Institute of Mathematics and Informatics, Vilnius, Lithuania.

Let y and z be two study variables defined on the population Consider a finite population : of N elements. Let y and z be two study variables defined on the population U and taking values and respectively.

- are sample design weights. We are interested in the estimation of the covariance Let us consider the standard estimator: - denotes a probability sample set. - are sample design weights. - is a probability of inclusion of the element k into the sample set s.

Let be known auxiliary variables with known covariance We construct three types of estimators of the covariance The weights are defined using three different calibration equations.

1. The nonlinear calibration The calibrated weights satisfy the calibration equation

2. The linear calibration The weights are defined by the calibration equation

3. Calibration of the total The calibration equations are

Different loss functions can be used: ♦ ♦ ♦ ♦ ♦ ♦

Let us introduce some notation:

Proposition 1. The weights , which satisfy the calibration equation and minimize the loss function can be expressed as with

Proposition 2. The weights , which satisfy the calibration equation and minimize the loss function can be expressed as with here is a properly chosen root of the equation

Proposition 3. The weights , which satisfy the calibration equation and minimize the loss function can be expressed as with

Proposition 4. The weights , which satisfy the calibration equation and minimize the loss function can be expressed as with

Simulation results We compare three types of calibrated estimators with known estimators of population covariance:

We have used for simulation distance functions and and three types of calibration equation. Six cases: I type calibration: and II type calibration: and III type calibration: and

 Case Bias MSE cv   -3490848 4.53E+13 0.0919 -3489778 4.52E+13 -6104197 9.48E+13 0.1265 -6111007 9.50E+13 0.1267 546136 2.93E+13 0.0808 535256 2.92E+13 0.0807 -5602394 1.55E+14 0.1836 -5455553 1.54E+14 0.1835 Case   Bias  MSE  cv   -4209064 1.03E+14 0.1493 -4216576 0.1494 -5069762 8.37E+13 0.1248 -5085344 8.40E+13 0.1250 -5874261 1.50E+14 0.1782 -5881074 -5082344 1.49E+14 0.1820 -4934908 1.48E+14

Case   Bias MSE   cv   -4690847 1.73E+14 0.2004 -4854206 1.70E+14 0.1978 -5562585 1.56E+14 0.1848 -5561796 0.1846 -5912708 1.60E+14 0.1855 -5911844 -5662564 1.57E+14 -5515768 0.1847

Correlation 0,8 Correlation 0,4 Case Bias MSE cv -3675 9.17E+08 0.0867   -3675 9.17E+08 0.0867 -3450 9.03E+08 0.0861 3671 1.43E+09 0.1063 3402 4484 7.32E+08 0.0752 4828 7.16E+08 0.0741 -125 1.58E+09 0.1134 306 Correlation 0,4 Case Bias MSE cv   2063 1.57E+09 0.1124 2427 1.53E+09 0.1108 2498 1.60E+09 0.1131 2292 1.59E+09 0.1130 20653 2.64E+09 0.1269 20720 2.58E+09 0.1251 -130 0.1127 341 0.1126

Correlation 0,9 Correlation 0,6 Case Bias MSE cv -345948 1.76E+11   -345948 1.76E+11 0.2779 -344035 1.74E+11 0.2761 -446667 3.85E+11 0.5741 -442287 3.82E+11 0.5721 -176199 5.03E+10 0.1360 -170965 4.83E+10 0.1345 -286681 5.08E+11 0.7174 -283628 5.09E+11 0.7172 Correlation 0,6 Case Bias MSE cv   -450255 3.86E+11 0.5744 -448187 3.85E+11 0.5740 -382883 4.87E+11 0.7170 -380149 4.85E+11 0.7150 -436150 3.22E+11 0.4770 -433248 3.19E+11 0.4758 -291121 5.14E+11 0.7300 -288024 5.15E+11 0.7299

Correlation 0,3 Case Bias MSE cv -444330 4.92E+11 0.7213 -443450   -444330 4.92E+11 0.7213 -443450 4.90E+11 0.7200 -402524 5.24E+11 0.7586 -401682 5.23E+11 0.7566 -473459 4.27E+11 0.6228 -471243 4.26E+11 0.6227 -293768 5.13E+11 0.7445 -281019 0.7443

Correlation 0,8 Correlation 0,4 Case Bias MSE cv -120894 1.36E+10 0.0738 -121548 -129208 1.77E+10 0.0845 -129470 1.78E+10 65118 1.05E+10 0.0643 65239 -118643 2.76E+10 0.1055 -66528 Correlation 0,4 Case Bias MSE cv -112093 2.50E+10 0.1004 -111965 -162762 2.40E+10 0.0984 -162495 -73870 2.83E+10 0.1067 -74058 2.82E+10 0.1066 -117655 2.69E+10 0.1042 -65692