Research Article
Combined Prediction Model of Death Toll for Road Traffic Accidents Based on Independent and Dependent Variables
Table 8
The combined model prediction value of traffic death tolls in China from 2002 to 2011.
| Year | Fatalities/person | Prediction value of combined model/person | Relative error/% | Relative error of Verhulst model/% | Relative error of multiple linear regression model/% |
| 2002 | 109381 | 109595 | 0.196 | 0 | 0.871 | 2003 | 104372 | 104384 | 0.011 | 0.188 | 0.700 | 2004 | 99217 | 98590 | 0.632 | 0.361 | 1.569 | 2005 | 98738 | 94190 | 4.606 | 5.337 | 2.083 | 2006 | 89455 | 88476 | 1.094 | 1.580 | 0.581 | 2007 | 81649 | 82714 | 1.304 | 1.190 | 1.695 | 2008 | 73484 | 76114 | 3.579 | 5.119 | 1.736 | 2009 | 67159 | 71478 | 6.431 | 7.141 | 3.977 | 2010 | 65225 | 66206 | 1.504 | 2.392 | 1.559 | 2011 | 62387 | 61829 | 0.894 | 0.991 | 0.564 | Comprehensive error | ā | ā | 2.025 | 2.700 | 1.534 |
|
|