Recognition of dominant driving factors behind sap flow of Liquidambar formosana based on back-propagation neural network method

Table 2 Fitting accuracy statistics of training and testing datasets from eight schemes

Data type		Scheme 1	Scheme 2	Scheme 3	Scheme 4	Scheme 5	Scheme 6	Scheme 7	Scheme 8
Max Acc	Training	0.8165	0.8173	0.7946	0.8877	0.8948	0.8983	0.9019	0.9032
Max Acc	Testing	0.7954	0.7830	0.7443	0.8636	0.8680	0.8657	0.8929	0.8956
Min Acc	Training	0.7679	0.7808	0.7488	0.8596	0.8533	0.8752	0.8752	0.8644
Min Acc	Testing	0.7470	0.7321	0.7038	0.8041	0.8025	0.8673	0.8370	0.8289
Med Acc	Training	0.776	0.8034	0.7698	0.8799	0.8817	0.8886	0.8731	0.8894
Med Acc	Testing	0.7633	0.7810	0.7218	0.8655	0.8306	0.8667	0.8278	0.8321
Ave Acc	Training	0.7618	0.8027	0.7687	0.8765	0.8794	0.8830	0.8788	0.8843
Ave Acc	Testing	0.7650	0.7880	0.7209	0.8519	0.8251	0.8417	0.8231	0.8299

ISSN: 1297-966X