Evaluation of the Urban Low-Carbon Sustainable Development Capability Based on the TOPSIS-BP Neural Network and Grey Relational Analysis

Table 4

Comparison of actual value and predicted value of the TOPSIS-BP neural network model.

Years	Expected value	Predicted value	Relative error (%)
Beijing
2009	0.1918	0.1918	0.0074
2010	0.1548	0.1548	0.0258
2011	0.1931	0.1931	0.0199
2012	0.2381	0.2383	0.0498
2013	0.6968	0.6968	0.0003
2014	0.2803	0.2805	0.0739
2015	0.2927	0.2925	0.0480
2016	0.3077	0.3075	0.0416
2017	0.3127	0.3126	0.0427
2018	0.3584	0.3584	0.0009
Shanghai
2009	0.3977	0.3982	0.1151
2010	0.3789	0.3794	0.1469
2011	0.3352	0.3354	0.0616
2012	0.3042	0.3043	0.0549
2013	0.3317	0.3320	0.1143
2014	0.4226	0.4216	0.2325
2015	0.4409	0.4410	0.0220
2016	0.5086	0.5097	0.2247
2017	0.5137	0.5137	0.0076
2018	0.5694	0.5695	0.0161
Guangzhou
2009	0.3190	0.3191	0.0399
2010	0.2885	0.2885	0.0075
2011	0.2085	0.2087	0.0970
2012	0.2008	0.2025	0.8326
2013	0.2274	0.2263	0.4907
2014	0.2450	0.2453	0.1252
2015	0.2632	0.2590	1.6083
2016	0.2621	0.2655	1.3161
2017	0.6150	0.6151	0.0240
2018	0.6978	0.6977	0.0277
Shenzhen
2009	0.3068	0.3057	0.3457
2010	0.2981	0.2994	0.4331
2011	0.3793	0.3790	0.0806
2012	0.4743	0.4743	0.0099
2013	0.3843	0.3849	0.1608
2014	0.4474	0.4473	0.0106
2015	0.4388	0.4389	0.0144
2016	0.4587	0.4585	0.0472
2017	0.5440	0.5441	0.0158
2018	0.5455	0.5455	0.0092