TITLE

Two�Stage Genetic Algorithm in the Analysis of Biochemical Relations

AUTHOR(S)
Asikgil, Baris; Erar, Aydin
PUB. DATE
September 2012
SOURCE
Turkiye Klinikleri Journal of Biostatistics;2012, Vol. 4 Issue 2, p49
SOURCE TYPE
Academic Journal
DOC. TYPE
Article
ABSTRACT
Objective: Nonlinear regression analysis is usually used in medical or biochemical areas in order to estimate unknown parameters. Ordinary least squares method can be considered for parameter estimation. However, efficient parameter estimates can not be obtained with the help of this method when errors are autocorrelated. In this paper, a method called two-stage genetic algorithm (TSGA) is proposed in order to obtain efficient parameter estimates. Material and Methods: Two-stage least squares (TSLS) method is preferred in order to overcome the autocorrelation problem. However, the method has some problems in the parameter estimation process if initial values of parameters are not known. Therefore, TSGA can be used in the presence of unknown initial values for nonlinear parameters. In this paper, two data sets taken from Faculty of Pharmacy have been examined on a preferred nonlinear model and it has been noticed that residuals are autocorrelated. Because of both unknown initial values of parameters and the autocorrelation problem among errors, TSGA method has been applied by using MATLAB 7.1. Results: The results obtained by using genetic algorithm (GA) and TSGA have been compared in view of the correlograms. p-values in the correlograms obtained from GA are <0.001 for all lags. This means that the autocorrelation problem exists among errors. However, p-values in the correlograms obtained from TSGA are greater than a=0.05 for all lags. This means that the autocorrelation problem is removed. Moreover, mean squared error values for TSGA (4.86 and 0.03) are smaller than mean squared error values for GA (11.28 and 0.04). Conclusion: It can be concluded that TSGA can give efficient parameter estimates in the presence of autocorrelated errors.
ACCESSION #
82030037

 

Related Articles

  • Prediction of Acidity Constants of Thiazolidine-4-carboxylic Acid Derivatives Using Ab Initio and Genetic Algorithm-Partial Least Squares. Niazi, Ali; Jameh Bozorghi, Saeed; Nori Shargh, Davood // Turkish Journal of Chemistry;2006, Vol. 30 Issue 5, p619 

    A quantitative structure-property relationship study is suggested for the prediction of the acidity constants of some thiazolidine-4-carboxylic acid derivatives in aqueous solution. Ab initio theory was used to calculate some quantum chemical descriptors, including electrostatic potentials and...

  • Modeling and Prediction of Tree Height--Diameter Relationships Using Spatial Autoregressive Models. Junfeng Lu; Lianjun Zhang // Forest Science;Jun2011, Vol. 57 Issue 3, p252 

    Three spatial autoregressive models were applied to model the tree height- diameter relationship, including spatial lag model (SLM), spatial error model (SEM), and spatial Durbin model (SDM), with ordinary least squares (OLS) as a benchmark. Five spatial weight matrices were used to evaluate the...

  • OTOKORELASYON DURUMUNDA EN KÜÇÜK KARELER TEKNİĞİNÄ°N ALTERNATÄ°FÄ° OTOREGRESYON TEKNÄ°KLERÄ° VE BÄ°R UYGULAMA. ALBAYRAK, Ali Sait // Suleyman Demirel University Journal of Faculty of Economics & Ad;2014, Vol. 19 Issue 1, p1 

    Ordinary least squares (OLS) regression analysis is based on several statistical assumptions. One key assumption is that the errors are independent of each other. However, with time series data, the OLS residuals usually are correlated over time. It is not desirable to use OLS analysis for time...

  • An Autocorrelation Term Method for Curve Fitting. Houston, Louis M. // ISRN Applied Mathematics;2013, p1 

    The least-squares method is the most popular method for fitting a polynomial curve to data. It is based on minimizing the total squared error between a polynomial model and the data. In this paper we develop a different approach that exploits the autocorrelation function. In particular, we use...

  • r − k Class estimator in the linear regression model with correlated errors. Üstündagˇ Şiray, Gülesen; Kaçıranlar, Selahattin; Sakallıoğlu, Sadullah // Statistical Papers;May2014, Vol. 55 Issue 2, p393 

    Autocorrelation in errors and multicollinearity among the regressors are serious problems in regression analysis. The aim of this paper is to examine multicollinearity and autocorrelation problems concurrently and to compare the r − k class estimator to the generalized least squares...

  • The effects of spatial autoregressive dependencies on inference in ordinary least squares: a geometric approach. Smith, Tony; Lee, Ka // Journal of Geographical Systems;Jan2012, Vol. 14 Issue 1, p91 

    There is a common belief that the presence of residual spatial autocorrelation in ordinary least squares (OLS) regression leads to inflated significance levels in beta coefficients and, in particular, inflated levels relative to the more efficient spatial error model (SEM). However, our...

  • Simple regression-based tests for spatial dependence. Born, Benjamin; Breitung, Jörg // Econometrics Journal;Jul2011, Vol. 14 Issue 2, p330 

    We propose simple and robust diagnostic tests for spatial error autocorrelation and spatial lag dependence. The idea is to reformulate the testing problem such that the outer product of gradients (OPG) variant of the LM test can be employed. Our versions of the tests are based on simple...

  • AN EVOLUTIONARY APPROACH FOR IMPUTING MISSING DATA IN TIME SERIES. FIGUEROA GARCÍA, JUAN CARLOS; KALENATIC, DUSKO; LÓPEZ BELLO, CESAR AMILCAR // Journal of Circuits, Systems & Computers;Feb2010, Vol. 19 Issue 1, p107 

    This paper presents a proposal based on an evolutionary algorithm for imputing missing observations in time series. A genetic algorithm based on the minimization of an error function derived from their autocorrelation function, mean, and variance is presented. All methodological aspects of the...

  • Miscellanea Locally optimal designs for errors-in-variables models. KONSTANTINOU, M.; DETTE, H. // Biometrika;Dec2015, Vol. 102 Issue 4, p951 

    We consider the construction of optimal designs for nonlinear regression models when there are measurement errors in the covariates. Corresponding approximate design theory is developed for maximum likelihood and least-squares estimation, with the latter leading to nonconcave optimization...

Share

Read the Article

Courtesy of THE LIBRARY OF VIRGINIA

Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics