Table 4: Experiment results for Australian dataset using our proposed approach and grid search.

Fold numberOur proposed approachGrid search method
Number of featuresNumber of support vectorsAccuracy (%)Number of featuresNumber of support vectorsAccuracy

1368.082991.48664517385.5072320.1251420282.6087
2827.071801.65084620691.30430.52.01432489.8551
3529.604390.32159720389.85510.51.01426186.7647
41845.72470.95494319894.28570.1250.031251449981.1594
5115.072624.63776621592.85715120.06251419386.9565
6347.313110.29031719789.855110240.031251419789.8551
7379.013916.52218321489.70590.52.01431988.2353
8131.805644.06696520490.00001280.1251418976.8116
9584.012935.78718322988.2353320.1251420582.8571
1074.4645883.41797618889.70590.1250.0156251450490.0000

Avg5.1202.790.131214289.385.5104