Advances in Multimedia

Research Article

Commercial Video Evaluation via Low-Level Feature Extraction and Selection

Table 8

Experiment results for video popularity prediction using different methods.
Feature numberCFS-SpearmanCFSmRMRp-norm (p=0.9)
Feature setCorrect classificat-ion rate (%)Feature setCorrect classificat-ion rate (%)Feature setCorrect classificat-ion rate (%)Feature setCorrect classificat-ion rate (%)
k=156.666756.666750.666756.6667
k=2(4,14)69.3333(5,14)59.3333(4,12)63.3333(5,14)59.3333
k=3(4,12,14)78(5,13-14)56.6667(4,12,16)76(4,12,14)64.6667
k=4(4,7,12,14)75.3333(5,13-15)68.6667(4,9,12,16)73.3333(4,9,12,14)72
k=5(2-5,7)77.3333(5,12-15)69.3333(4,6,9,12,16)74.6667(4,5,9,12,14)74.3333
k=6(2-5,7,14)76(5,12-16)75.3333(2,4,6,9,12,16)74.6667(2,4,5,9,12,14)74.6667
k=7(1-5,7,14)76(5-6,12-16)74.6667(2,4,6,9,12,15-16)76(1,2,4,5,9,12,14)76
k=8(1-5,7-8,14)74.6667(5-6,10,12-14)74.6667(2,4-6,9,12,15-16)76(1,2,4-6,9,12,14)74.6667
k=9(1-8,14)74.6667(5-6,10-16)74.6667(2-6,9,12,15-16)74.6667(1,2,4-6,9,12,14-15)74.6667
k=10(1-8,13-14)74.6667(5-6,9-16)74.6667(2-6,9-10,12,15-16)74.6667(1-6,9,12,14-15)74.6667
k=11(1-8,13-15)74.6667(4-6,9-16)74.6667(2-7,9-10,12,15-16)74.6667(1-6,9,12,14-16)72
k=12(1-8,12-15)74.6667(1,4-6,9-16)74.6667(2-7,9-12,15-16)74.6667(1-6,8,9,12,14-16)72
k=13(1-8,12-16)76(1,4-7,9-16)74.6667(1-7,9-12,15-16)74.6667(1-9,12,14-16)72
k=14(1-8,10-14,16)72(1,4-16)76(1-12,15-16)74(1-9,11-12,14-16)71.6667
k=15(1-8,10-16)75.3333(1,3-16)75.3333(1-12,14-16)75.3333(1-12,14-16)75.3333