Research Article
River Flow Estimation from Upstream Flow Records Using Support Vector Machines
Table 4
Average SSE values of the SVM approach for Wilson data, Viessman and Lewis data, and River Wyre data.
| Data | | | | | |
| Wilson | 0.9 | 1.389 | 1.627 | 2.013 | 2.691 | 0.8 | 3.486 | 4.120 | 4.160 | 7.738 | 0.7 | 7.284 | 8.473 | 8.835 | 10.818 | 0.6 | 12.540 | 14.872 | 17.878 | 21.755 |
| Viessman and Lewis | 0.9 | 12460.720 | 14446.716 | 18456.445 | 21172.915 | 0.8 | 34631.312 | 48888.423 | 50163.966 | 58837.181 | 0.7 | 83005.647 | 84387.246 | 112145.221 | 100247.301 | 0.6 | 124339.197 | 124324.518 | 118768.183 | 131235.761 |
| River Wyre | 0.9 | 5.760 | 9.895 | 11.922 | 7.788 | 0.8 | 218.236 | 55.846 | 70.096 | 289.540 | 0.7 | 291.692 | 161.367 | 133.620 | 363.987 | 0.6 | 331.883 | 271.468 | 239.053 | 440.319 |
|
|