This is based on simple linear regressor on multiple items inthe dataset for simple regressor we use the formula called,y=b0+b1x1 but for yhe multiple linear regressor we extend the formula by,y=b0+b1x1+b2x2+b3x3+...+bn*xn We use this by usage of the simple backward elimiination of the Significant errors and it is smaller to bigger to approach.