skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Maximum likelihood method for parameter estimation in linear model with below-detection data

Abstract

The maximum likelihood (ML) method for regression analyses of left-censored data is improved for general acceptance by considering the censored observations to be doubly censored. The existence of a lower bound (i.e., the concentration of a pollutant cannot be negative) is included; the improved ML method utilizes this information in the formulation of a likelihood function. The improved ML method has been translated into an equivalent least squares (LS) method, and an iterative algorithm is presented to estimate the statistical parameters from this LS translation. The LS translation is easy to explain to nonstatisticians, and computational requirements for implementing the LS method are minimal. The methodology is applied to a mechanistic model for air transport and deposition of polycyclic aromatic hydrocarbons (PAH) to a snow surface. For a censored data set, parameter estimates of the model, namely, dry deposition velocities and washout ratios, were obtained for various PAH species by using the following three procedures: (1) the NAG-15 routine for maximization of a likelihood function; (2) the proposed algorithm for the equivalent LS method; and (3) the modified iterative least squares method.

Authors:
;  [1];  [2]
  1. Univ. of Waterloo, Ontario (Canada). Dept. of Civil Engineering
  2. CRA, Ltd., Waterloo, Ontario (Canada)
Publication Date:
Sponsoring Org.:
USDOE
OSTI Identifier:
162994
Resource Type:
Journal Article
Resource Relation:
Journal Name: Journal of Environmental Engineering; Journal Volume: 121; Journal Issue: 11; Other Information: PBD: Nov 1995
Country of Publication:
United States
Language:
English
Subject:
54 ENVIRONMENTAL SCIENCES; POLYCYCLIC AROMATIC HYDROCARBONS; ENVIRONMENTAL TRANSPORT; ALGORITHMS; LEAST SQUARE FIT; WASHOUT; AIR POLLUTION

Citation Formats

Sharma, M., Thomson, N., and McBean, E.A. Maximum likelihood method for parameter estimation in linear model with below-detection data. United States: N. p., 1995. Web. doi:10.1061/(ASCE)0733-9372(1995)121:11(776).
Sharma, M., Thomson, N., & McBean, E.A. Maximum likelihood method for parameter estimation in linear model with below-detection data. United States. doi:10.1061/(ASCE)0733-9372(1995)121:11(776).
Sharma, M., Thomson, N., and McBean, E.A. 1995. "Maximum likelihood method for parameter estimation in linear model with below-detection data". United States. doi:10.1061/(ASCE)0733-9372(1995)121:11(776).
@article{osti_162994,
title = {Maximum likelihood method for parameter estimation in linear model with below-detection data},
author = {Sharma, M. and Thomson, N. and McBean, E.A.},
abstractNote = {The maximum likelihood (ML) method for regression analyses of left-censored data is improved for general acceptance by considering the censored observations to be doubly censored. The existence of a lower bound (i.e., the concentration of a pollutant cannot be negative) is included; the improved ML method utilizes this information in the formulation of a likelihood function. The improved ML method has been translated into an equivalent least squares (LS) method, and an iterative algorithm is presented to estimate the statistical parameters from this LS translation. The LS translation is easy to explain to nonstatisticians, and computational requirements for implementing the LS method are minimal. The methodology is applied to a mechanistic model for air transport and deposition of polycyclic aromatic hydrocarbons (PAH) to a snow surface. For a censored data set, parameter estimates of the model, namely, dry deposition velocities and washout ratios, were obtained for various PAH species by using the following three procedures: (1) the NAG-15 routine for maximization of a likelihood function; (2) the proposed algorithm for the equivalent LS method; and (3) the modified iterative least squares method.},
doi = {10.1061/(ASCE)0733-9372(1995)121:11(776)},
journal = {Journal of Environmental Engineering},
number = 11,
volume = 121,
place = {United States},
year = 1995,
month =
}
  • We demonstrate in this report an envelope detection technique with maximum likelihood estimation in a least square sense for determining displacement. This technique is achieved by sampling the amplitudes of quadrature signals resulted from a heterodyne interferometer so that the resolution of displacement measurement of the order of {lambda}/10{sup 4} is experimentally verified. A phase unwrapping procedure is also described and experimentally demonstrated and indicates that the unambiguity range of displacement can be measured beyond a single wavelength.
  • In this study, we derive a comprehensive forward model for the data collected by stripmap synthetic aperture radar (SAR) that is linear in the ground reflectivity parameters. It is also shown that if the noise model is additive, then the forward model fits into the linear statistical model framework, and the ground reflectivity parameters can be estimated by statistical methods. We derive the maximum likelihood (ML) estimates for the ground reflectivity parameters in the case of additive white Gaussian noise. Furthermore, we show that obtaining the ML estimates of the ground reflectivity requires two steps. The first step amounts tomore » a cross-correlation of the data with a model of the data acquisition parameters, and it is shown that this step has essentially the same processing as the so-called convolution back-projection algorithm. The second step is a complete system inversion that is capable of mitigating the sidelobes of the spatially variant impulse responses remaining after the correlation processing. We also state the Cramer-Rao lower bound (CRLB) for the ML ground reflectivity estimates.We show that the CRLB is linked to the SAR system parameters, the flight path of the SAR sensor, and the image reconstruction grid.We demonstrate the ML image formation and the CRLB bound for synthetically generated data.« less
  • Restricted Maximum Likelihood (REML) estimation of variance-covariance matrices is an optimization procedure that behaves well in the presence of sampling bias and that has both scientific and industrial applications. Parallel algorithms are presented and compared for computing the REML gradient when the covariance matrix is not assumed block diagonal. The implementations presented are based on the Portable Extensible Toolkit for Scientific Computing (PETSc) and can run on any parallel computer supporting MPI.
  • The method of maximum likelihood is applied to the estimation of the magnetic curvature of charged particle trajectories in bubbIe chambers, using measuremeats made on points approximately equidistant from the projection of the tracks on a plane perpendicular to the magnetic field. (tr-auth)
  • The method of maximum likelihood is used in the estimntion of the angular coefficient of the tangent to a charged particle trajectory in a bubble chamber, given the coordinates of approximately equidistant points. A weighted average, whose weights are calculated as functions of parameter expressing the relative importance of multiple scattering and measurement errors, is thus obtained as an estimator. The variance of this estimator is calculated and compared with the variances of very simple weighted average s. (tr -auth)