Page - 353 - in Book of Full Papers - Symposium Hydro Engineering

4. MULTIPLE REGRESSIVE FORMULA According to the analysis above, using forward and stepwise regressive methods, we got a full regression model (considering all independent variables) and a selecting regression model (Selecting most important independent variables) separately as follows: ?̂? 1 = 0.265?̂? 2 ∗0.417?̂? 3 0.415?̂? 3 ∗−0.070?̂? 4 −0.031?̂? 5 −0.030?̂? 6 0.951 (R2=0.642) [4] ?̂? 1 = 0.0736?̂? 1 0.415?̂? 6 1.151 (R2=0.631) [5] ?̂? 2 = 185.353?̂? 2 ∗1.373?̂? 3 −0.430?̂? 3 ∗0.248?̂? 4 0.317?̂? 5 0.320?̂? 6 0.719 (R2=0.827) [6] ?̂? 2 = 3.540?̂? 2 ∗1.428 (R2=0.801) [7] Figure 2 and figure 3 gave the results of the regression models. (a) full regression model (b) selecting regression model Figure 2. Multiple regressive result of height increment of sand bars in the tributaries in Xiaolangdi Reservoir Area (a) full regression model (b) selecting regression model Figure 3. Multiple regressive result of annul sedimentation quantity in the tributaries in Xiaolangdi Reservoir Area 353