Research Article

The Influence of Data Length on the Performance of Artificial Intelligence Models in Predicting Air Pollution

Table 3

The performance of the proposed models through the training phase.

MAE ()
Length of data %GBR-M1ELM-M1GMDHNN-M1GBR-M2ELM-M2GMDHNN-M2GBR-M3ELM-M3GMDHNN-M3

507.27991.10991.11877.44881.08891.11346.86591.08281.1131
553.70641.09371.10087.20831.07381.09667.58511.06431.0936
607.52981.07211.07727.52521.05251.07295.92511.04801.0711
657.42811.05011.05564.20151.02891.05157.42391.02601.0484
706.63361.03171.03656.98241.01331.03334.92801.00631.0293
757.50041.02061.02584.23291.00401.02136.84011.00851.0178
807.48741.00711.01177.48510.98951.00746.03590.98521.0038
856.26300.99190.99547.39710.98180.99857.39120.97670.9936
907.36700.99160.99557.36360.97600.99107.36070.97100.9898
Min3.70640.99160.99544.20150.97600.99104.92800.97100.9898
Max7.52981.10991.11877.52521.08891.11347.58511.08281.1131

R
500.90630.90620.90570.91110.91060.90710.91000.91370.9076
550.90510.90630.90600.90920.91110.90740.91080.91450.9084
600.90720.90670.90640.91450.91110.90770.90830.91330.9087
650.90820.90760.90730.90040.91310.90870.91610.91430.9096
700.90840.90850.90820.91150.91260.90940.90980.91580.9103
750.91000.90950.90920.91000.91350.91080.91330.91340.9116
800.90990.91020.90990.91690.91410.91130.91230.91640.9121
850.90940.91100.91070.91560.91460.91100.91640.91620.9119
900.91130.91100.91070.91700.91510.91230.91680.91710.9127
Min0.90510.90620.90570.90040.91060.90710.90830.91330.9076
Max0.91130.91100.91070.91700.91510.91230.91680.91710.9127

RMSE ()
509.41151.83291.83729.60291.79131.82448.85181.76151.8197
554.85361.81021.81389.27541.76611.80039.78941.73341.7916
609.67671.77861.78179.69991.73781.76947.61491.71791.7601
659.54951.74651.74935.47721.69631.73709.56441.68531.7288
708.46491.71601.71868.88481.67921.70776.30411.64881.6994
759.55551.70321.70595.43471.66741.69198.68651.66761.6847
809.52711.68771.69049.52541.65181.67757.67361.63071.6703
858.00181.66481.66719.42561.63601.66839.41381.62091.6602
909.39971.66481.66719.40561.62751.65259.38571.60881.6495
Min4.85361.66481.66715.43471.62751.65256.30411.60881.6495
Max9.67671.83291.83729.69991.79131.82449.78941.76151.8197

WI
500.64630.94890.94850.64130.95150.94950.66500.95330.9498
550.81900.94900.94880.64730.95180.94960.63250.95380.9502
600.62910.94920.94900.63130.95180.94980.69930.95310.9503
650.62960.94980.94960.78130.95300.95050.63230.95370.9507
700.66030.95030.95020.64640.95270.95090.74420.95460.9515
750.62380.95090.95080.78200.95320.95160.65270.95320.9520
800.62280.95130.95100.62530.95360.95190.68700.95490.9524
850.67220.95170.95160.62620.95380.95180.62660.95480.9522
900.62520.95170.95160.62710.95410.95250.62740.95530.9528
Min0.62280.94890.94850.62530.95150.94950.62660.95310.9498
Max0.81900.95170.95160.78200.95410.95250.74420.95530.9528

NSE
500.61240.94090.94040.60340.94200.94070.63440.94230.9407
550.79810.94040.94000.60730.94150.94030.58670.94200.9404
600.57650.93970.93940.57680.94080.93970.66680.94110.9398
650.57090.93930.93900.75730.94060.93930.57110.94070.9394
700.60660.93880.93850.58590.93990.93870.70770.94030.9390
750.55330.93920.93890.74790.94020.93920.59260.93990.9394
800.54900.93930.93910.54920.94040.93930.63650.94070.9395
850.61550.93910.93890.54800.94000.93900.54840.94030.9393
900.54770.93910.93890.54790.94010.93920.54810.94040.9392
Min0.54770.93880.93850.54790.93990.93870.54810.93990.939
Max0.79810.94090.94040.75730.9420.94070.70770.94230.9407

Bold values indicate the highest accuracy.