Dynamic Heat Supply Prediction Using Support Vector Regression Optimized by Particle Swarm Optimization Algorithm

Table 3

Comparison of the experimental results obtained with the PSO-SVR model and six data sets.

Samples	Size	Parameters			Evaluation indexes
					Training set MSE	Training set	Testing set MSE	Testing set
Data Set 1	196	99.2447	0.022	7.75 × 10^−5	0.0019	0.9945		0.9475
Data Set 2	248	98.8186	0.0096	2.24 × 10^−4	0.0022	0.9907		0.9798
Data Set 3	240	95.4227	0.0061	3.23 × 10^−4	0.002	0.9878		0.9244
Data Set 4	176	95.9658	0.0006	2.27 × 10^−4	0.0025	0.9896		0.9887
Data Set 5	123	97.2685	0.0146	8.20 × 10^−5	0.0026	0.9896		0.9983
Data Set 6	104	96.8596	0.0152	3.56 × 10^−4	0.0027	0.9915		0.9605