Classification of imbalanced data is an important issue. Many algorithms have been developed for classification, such as Back Propagation (BP) neural networks, decision tree, Bayesian networks etc., and have been used repeatedly in many fields. These algorithms speak of the problem of imbalanced data, where there are situations that belong to more classes than others. Imbalanced data result in poor performance and bias to a class without other classes. In this paper, we proposed three techniques based on the Over-Sampling (O.S.) technique for processing imbalanced dataset and redistributing it and converting it into balanced dataset. These techniques are (Improved Synthetic Minority Over-Sampling Technique (Improved SMOTE),  Border

Because of the experience of the mixture problem of high correlation and the existence of linear MultiCollinearity between the explanatory variables, because of the constraint of the unit and the interactions between them in the model, which increases the existence of links between the explanatory variables and this is illustrated by the variance inflation vector (VIF), L-Pseudo component to reduce the bond between the components of the mixture.

    To estimate the parameters of the mixture model, we used in our research the use of methods that increase bias and reduce variance, such as the Ridge Regression Method and the Least Absolute Shrinkage and Selection Operator (LASSO) method a

This manuscript presents several applications for solving special kinds of ordinary and partial differential equations using iteration methods such as Adomian decomposition method (ADM), Variation iterative method (VIM) and Taylor series method. These methods can be applied as well as to solve nonperturbed problems and 3rd order parabolic PDEs with variable coefficient. Moreover, we compare the results using ADM, VIM and Taylor series method. These methods are a commination of the two initial conditions.

EDIRKTO, an Implicit Type Runge-Kutta  Method of Diagonally Embedded pairs, is a novel approach presented in the paper that may be used to solve 4th-order ordinary differential equations of the form . There are two pairs of EDIRKTO, with three stages each: EDIRKTO4(3) and EDIRKTO5(4). The derivation techniques of the method indicate that the higher-order pair is more accurate, while the lower-order pair provides superior error estimates. Next, using these pairs as a basis, we developed variable step codes and applied them to a series of -order ODE problems. The numerical outcomes demonstrated how much more effective their approach is in reducing the quantity of function evaluations needed to resolve fourth-order ODE issues.

The logistic regression model regarded as the important regression Models ,where of the most interesting subjects in recent studies due to taking character more advanced in the process of statistical analysis .                                                

The ordinary estimating methods is failed in dealing with data that consist of the presence of outlier values and hence on the absence of such that have undesirable effect on the result.    &nbs

Direct FIA methods for estimation of Naringenin (NG) in different samples were proposed. These methods are based on diazotization coupling reactions of two reagents: method (A) p-chloroaniline (PCA) and method (B) procaine hydrochloride (PRH) with NG in basic medium. Yellow dyes with maximum absorption at 416 and 415 nm were formed respectively. Calibration curves were constructed over different NG concentrations, linearity was from 1- 70 and 1- 40 µgmL^-1 with detection limits of 0.55 and 0.24 µgmL^-1 for (A) and (B) respectively. All analytical variables involved in the FIA procedure were evaluated and optimized. The established methods were successfully applied for the determination of NG in its supplements and u