Table 1.

Shows Comparison of MLP and SVM Algorithm

Column1	MlP (two hidden layers)	SVM
	The percentage split 75%	The percentage split 75%
Correlation coefficient	0.5669	0.6342
Mean absolute error	13.2958	11.6237
Root mean squared error	17.9234	17.1052
Relative absolute error	77.36%	67.63%
Root relative squared error	84.88%	81.00%
Total Number of Instances	359	359
	Percentage split 85%	Percentage split 85%
Correlation coefficient	0.6133	0.6402
Mean absolute error	14.372	11.2879
Root mean squared error	18.9334	17.0521
Relative absolute error	83.60%	65.66%
Root relative squared error	89.16%	80.30%
Total Number of Instances	216	216
Cross-validation (10 fold)
Correlation coefficient	0.5995	0.6419
Mean absolute error	12.7017	11.2861
Root mean squared error	17.1316	16.4005
Relative absolute error	75.31%	66.92%
Root relative squared error	83.20%	79.65%
Total Number of Instances	1438	1438
Cross-validation (15fold)
Correlation coefficient	0.61	0.6415
Mean absolute error	12.4953	11.2775
Root mean squared error	16.8053	16.4056
Relative absolute error	74.10%	66.88%
Root relative squared error	81.62%	79.68%
Total Number of Instances	1438	1438