An approach of the grey modeling based on ax + b,sin x over x transformation

This paper point out the method of ax + b and sin x/ x function transformation against the modeling data. Proving that ax + b( 0 <; a <;1, b> or -1<; a <; 0, b > 0, x<; -b/2a), sin x/ x (2 kπ <; x <; 2 kπ + π/2) could improve the smoothness of data, reduces the information difference and class ratio dispersion, ax + b is more effective than x + b when 0 <; a <; 1,b > or -1<; a <; 0, b > 0, x<;-b(a+1)/2a, sin x / x is more effective than cos x when 2 kπ <; × <; 2 kπ + π/4.