Use of BP Neural Networks to Determine China’s Regional CO₂ Emission Quota

Table 8

Comparison between model results and historical quotas.

Year	Region	Historical quotas	Benchmark model		New model
			Training results	Loss	Training results	Loss
2014	Beijing	0.5	0.5114	0.0114	0.502	0.002
	Tianjing	1.6	1.7397	0.1397	1.584	−0.016
	Shanghai	1.5	1.7795	0.2795	1.4785	−0.0215
	Hubei	3.24	3.0159	−0.2241	3.2485	0.0085
	Guangdong	3.88	3.6819	−0.1981	3.8754	−0.0046
	Shenzhen	0.33	0.3623	0.0323	0.3169	−0.0131
	Chongqing	1.3	1.3283	0.0283	1.2949	−0.0051
2015	Beijing	0.5	0.5441	0.0441	0.5144	0.0144
	Tianjing	1.6	1.7312	0.1312	1.6169	0.0169
	Shanghai	1.6	1.6728	0.0728	1.6214	0.0214
	Hubei	3.24	3.0162	−0.2238	3.1589	−0.0811
	Guangdong	4.08	3.8977	−0.1823	4.0595	−0.0205
	Shenzhen	0.33	0.3623	0.0323	0.3182	−0.0118
	Chongqing	1.25	1.419	0.169	1.3037	0.0537
2016	Beijing	0.5	0.4536	−0.0464	0.4882	−0.0118
	Tianjing	1.6	1.5162	−0.0838	1.5506	−0.0494
	Shanghai	1.5	1.4923	−0.0077	1.5099	0.0099
	Hubei	2.8	2.9384	0.1384	2.8715	0.0715
	Guangdong	4	4.0552	0.0552	4.0141	0.0141
	Shenzhen	0.3	0.3636	0.0636	0.3209	0.0209
	Chongqing	1.3	1.4176	0.1176	1.2564	−0.0436
2017	Beijing	0.5	0.3623	−0.1377	0.4946	−0.0054
	Tianjing	1.6	1.363	−0.237	1.6399	0.0399
	Shanghai	1.6	1.2582	−0.3418	1.5741	−0.0259
	Hubei	2.5	2.5891	0.0891	2.4883	−0.0117
	Guangdong	4.2	4.5011	0.3011	4.208	0.008
	Shenzhen	0.3	0.3623	0.0623	0.3043	0.0043
	Chongqing	1.3	1.2166	−0.0834	1.2883	−0.0117

Unit of quotas: 100 million tons.