The estimation of the parameters of linear regression is based on the usual Least Square method, as this method is based on the estimation of several basic assumptions. Therefore, the accuracy of estimating the parameters of the model depends on the validity of these hypotheses. The most successful technique was the robust estimation method which is minimizing maximum likelihood estimator (MM-estimator) that proved its efficiency in this purpose. However, the use of the model becomes unrealistic and one of these assumptions is the uniformity of the variance and the normal distribution of the error. These assumptions are not achievable in the case of studying a specific problem that may include complex data of more than one model. To deal with this type of problem, a mixture of linear regression is used to model such data. In this article, we propose a genetic algorithm-based method combined with (MM-estimator), which is called in this article (RobGA), to improve the accuracy of the estimation in the final stage. We compare the suggested method with robust bi-square (MixBi) in terms of their application to real data representing blood sample. The results showed that RobGA is more efficient in estimating the parameters of the model than the MixBi method with respect to mean square error (MSE) and classification error (CE).
The current research aims to know the complementary use of the paper book and the digital book by defining the paper book and the digital book and knowing the methods and methods of obtaining each of them and indicating the extent of their use and for what purpose and knowing the difficulties that researchers face in using each of them and which is better using the paper book or the digital book in preparing research Studies and for no reason. The research relied on the descriptive approach and the questionnaire as a tool for data collection, and it was distributed to professors and students of Iraqi universities. The research sample reached (219) individuals and reached a set of results, the most important of which are: Preference for the
... Show MoreThis paper is concerned with introducing and studying the new approximation operators based on a finite family of d. g. 'swhich are the core concept in this paper. In addition, we study generalization of some Pawlak's concepts and we offer generalize the definition of accuracy measure of approximations by using a finite family of d. g. 's.
In petroleum industry, the early knowledge of “pore pressure gradient” is the basis in well design and the extraction of these information is more direct when the pore pressure gradient is equal to normal gradient; however, this matter will be more complex if it deviate from that limit which is called “abnormal pore pressure”, if this variable does not put in consideration, then many drilling problems will occur might lead to entire hole loss. To estimate the pore pressure gradient there are several methods, in this study; Eaton method’s is selected to extract the underground pressure program using drilling data (normalized rate of penetration) and logs data (sonic and density log). The results shows that an abnormal high press
... Show MoreVariable selection in Poisson regression with high dimensional data has been widely used in recent years. we proposed in this paper using a penalty function that depends on a function named a penalty. An Atan estimator was compared with Lasso and adaptive lasso. A simulation and application show that an Atan estimator has the advantage in the estimation of coefficient and variables selection.
This paper is interested in comparing the performance of the traditional methods to estimate parameter of exponential distribution (Maximum Likelihood Estimator, Uniformly Minimum Variance Unbiased Estimator) and the Bayes Estimator in the case of data to meet the requirement of exponential distribution and in the case away from the distribution due to the presence of outliers (contaminated values). Through the employment of simulation (Monte Carlo method) and the adoption of the mean square error (MSE) as criterion of statistical comparison between the performance of the three estimators for different sample sizes ranged between small, medium and large (n=5,10,25,50,100) and different cases (wit
... Show MoreIn this paper, wavelets were used to study the multivariate fractional Brownian motion through the deviations of the random process to find an efficient estimation of Hurst exponent. The results of simulations experiments were shown that the performance of the proposed estimator was efficient. The estimation process was made by taking advantage of the detail coefficients stationarity from the wavelet transform, as the variance of this coefficient showed the power-low behavior. We use two wavelet filters (Haar and db5) to manage minimizing the mean square error of the model.
Throughout this paper we study the properties of the composition operator
C
p1 o
p2 o…o
pn induced by the composition of finite numbers of special
automorphisms of U,
pi (z) i
i
p z
1 p z
Such that pi U, i 1, 2, …, n, and discuss the relation between the product of
finite numbers of automorphic composition operators on Hardy space H2 and some
classes of operators.
In this paper, some estimators for the unknown shape parameter and reliability function of Basic Gompertz distribution have been obtained, such as Maximum likelihood estimator and Bayesian estimators under Precautionary loss function using Gamma prior and Jefferys prior. Monte-Carlo simulation is conducted to compare mean squared errors (MSE) for all these estimators for the shape parameter and integrated mean squared error (IMSE's) for comparing the performance of the Reliability estimators. Finally, the discussion is provided to illustrate the results that summarized in tables.