polynomial regression

This paper introduces a new proposed algorithm of numerical integration evaluation regarded as optimization problem solution. The new method is characterized to have superiority features such as attractive, accurate and rapid. An improvement of polynomial regression has been done by selecting nearest neighbors points being searched around of the values of regression coefficients which calculated by using least squares method. Furthermore, Trapezoidal and Simpson methods were considered as traditional methods in numerical integration. In this regard, comparison has been done among all four methods used in simulation application via MATLAB program that have been performed to achieve the desired numerical results for the four methods. As conclusion, the proposed algorithm approved its superiority.