说明：  曲线拟合：通过给定的一组二维坐标点，利用线性代数中矩阵求逆的知识，从这些数据中找出规律性，预测原函数f（x)的表达式，以消除所给数据的局部波动。一次拟合出来的函数图像是直线，二次拟合出来的图像是抛物线。两者相比，二次拟合逼近程度更高。
(curve fitting through a given set of two-dimensional coordinates, using the inverse matrix in linear algebra knowledge, find out the rules from these data, the original prediction function f (x) expression, to eliminate the fluctuation to local data. Once the function image is a straight line, two times fitting the image is parabolic. The two fitting approximation is higher than the latter.)