using backpropagation neural network for fuel consumption prediction
I want to use three datasets in the model development; the worked hours, the loads and the fuel consumed by the truck.I want to use the worked hours and loads as the inputs(Two Inputs) and the fuel consumed as the target(output data). my dataset is consist of 80 records and i want to divide it into two sub datasets:training and testing. 65% of the total data for training and the remaining 35% for testing so as to validate accuracy of the model.The number of neuron in the hidden layer should be set to 1 initially and the hyperbolic tangent sigmiod function should be used as the transfer function.The data should be run to be run to give an output with the calculated error between the target and the output(predicted). The error should then be back propagated through the network for adjustments.The process should be repeated by varing the neuron number till the error is minimal. The model should then be simulated to ascertain the validity by using the testing data.The optimim number of neuron that gives the best results should be selected based on the correlation coefficient and the mean absolute percentage error for the testing data.The architecture of the BPNN should be developed.Find the data to be used in is below
WORKED HOURS(Hrs) : 299.9 372.3 401.9 429.5 358.3 79.6 310.3 176.4 221.05 187.6 433.6 243 480.5 203.5 426.8 456.6 473.2 350.4 474 161.4 390.8 471.7 266.3 496.6 423.7 320.9 18.1 504.5 531.5 514.8 519.3 513.2 49 206.3 416.8 290.9 368.9 484.6 209 225.6 439.6 461.9 352.3 455.3 475.6 445.1 346.5 469.7 288.7 306.3 170.6 391.7 147.4 392 343.6 223.6 67.1 361.6 279.5 361.3 451.7 76.6 344.8 352.3 347.7 450.6 434.7 365.9 327.6 451.5 433 361.9 180.3 466.5 415.6 461.7 437.1 478.3 325.2 499.7
LOADS (BCM) 1129 972 703 1199 1438 1144 1092 1526 772 1018 1415 1026 424 931 489 1044 135 181 1488 1038 1112 1191 149 1607 830 944 595 744 1126 1386 922 1217 1471 1124 1490 1436 1095 1266 1245 1384 1183 1299 940 1215 1098 1125 929 1307 1253 1407 1471 836 994 1138 1576 1066 1138 1106 1170 1283 942 1420 663 1167 1118 1254 1214 1187 1387 1311 1108 1336 1082 1435 1326 1188 1527 1036 1172 1517
FUEL CONSUMED (l/bcm)
40,733.00 46,375.00 50,797.00 49,706.00 45,092.00 34,543.87 36,687.62 44,899.04 37,023.81 40,458.24 34,383.52 25,914.49 34,300.42 46,685.87 54,122.83 47,242.32 46,296.99 61,800.78 58,181.10 40,579.12 56,890.57 60,295.81 61,843.33 63,156.89 60,320.82 56,141.04 58,063.53 46,379.00 44,011.00 44,191.00 42,089.00 45,682.00 50,666.00 45,643.00 46,568.00 49,162.00 47,871.63 45,012.70 40,345.86 43,233.45 41,408.20 52,169.66 53,161.51 46,553.05 52,353.93 49,019.43 42,015.50 35,327.23 35,025.50 28,207.78 39,435.27 42,200.30 26,677.27 45,977.46 27,870.61 22,335.77 41,153.24 36,260.02 40,901.63 42,274.00 41,501.00 39,189.00 36,349.00 38,545.00 38,625.00 32,907.00 30,153.00 29,142.58 26,813.88 34,488.11 40,575.67 43,439.41 45,374.04 47,730.75 58,335.01 48,531.21 45,510.96 51,891.60 42,902.32 32,819.21I want to use three datasets in the model development; the worked hours, the loads and the fuel consumed by the truck.I want to use the worked hours and loads as the inputs(Two Inputs) and the fuel consumed as the target(output data). my dataset is consist of 80 records and i want to divide it into two sub datasets:training and testing. 65% of the total data for training and the remaining 35% for testing so as to validate accuracy of the model.The number of neuron in the hidden layer should be set to 1 initially and the hyperbolic tangent sigmiod function should be used as the transfer function.The data should be run to be run to give an output with the calculated error between the target and the output(predicted). The error should then be back propagated through the network for adjustments.The process should be repeated by varing the neuron number till the error is minimal. The model should then be simulated to ascertain the validity by using the testing data.The optimim number of neuron that gives the best results should be selected based on the correlation coefficient and the mean absolute percentage error for the testing data.The architecture of the BPNN should be developed.Find the data to be used in is below
WORKED HOURS(Hrs) : 299.9 372.3 401.9 429.5 358.3 79.6 310.3 176.4 221.05 187.6 433.6 243 480.5 203.5 426.8 456.6 473.2 350.4 474 161.4 390.8 471.7 266.3 496.6 423.7 320.9 18.1 504.5 531.5 514.8 519.3 513.2 49 206.3 416.8 290.9 368.9 484.6 209 225.6 439.6 461.9 352.3 455.3 475.6 445.1 346.5 469.7 288.7 306.3 170.6 391.7 147.4 392 343.6 223.6 67.1 361.6 279.5 361.3 451.7 76.6 344.8 352.3 347.7 450.6 434.7 365.9 327.6 451.5 433 361.9 180.3 466.5 415.6 461.7 437.1 478.3 325.2 499.7
LOADS (BCM) 1129 972 703 1199 1438 1144 1092 1526 772 1018 1415 1026 424 931 489 1044 135 181 1488 1038 1112 1191 149 1607 830 944 595 744 1126 1386 922 1217 1471 1124 1490 1436 1095 1266 1245 1384 1183 1299 940 1215 1098 1125 929 1307 1253 1407 1471 836 994 1138 1576 1066 1138 1106 1170 1283 942 1420 663 1167 1118 1254 1214 1187 1387 1311 1108 1336 1082 1435 1326 1188 1527 1036 1172 1517
FUEL CONSUMED (l/bcm)
40,733.00 46,375.00 50,797.00 49,706.00 45,092.00 34,543.87 36,687.62 44,899.04 37,023.81 40,458.24 34,383.52 25,914.49 34,300.42 46,685.87 54,122.83 47,242.32 46,296.99 61,800.78 58,181.10 40,579.12 56,890.57 60,295.81 61,843.33 63,156.89 60,320.82 56,141.04 58,063.53 46,379.00 44,011.00 44,191.00 42,089.00 45,682.00 50,666.00 45,643.00 46,568.00 49,162.00 47,871.63 45,012.70 40,345.86 43,233.45 41,408.20 52,169.66 53,161.51 46,553.05 52,353.93 49,019.43 42,015.50 35,327.23 35,025.50 28,207.78 39,435.27 42,200.30 26,677.27 45,977.46 27,870.61 22,335.77 41,153.24 36,260.02 40,901.63 42,274.00 41,501.00 39,189.00 36,349.00 38,545.00 38,625.00 32,907.00 30,153.00 29,142.58 26,813.88 34,488.11 40,575.67 43,439.41 45,374.04 47,730.75 58,335.01 48,531.21 45,510.96 51,891.60 42,902.32 32,819.21 I want to use three datasets in the model development; the worked hours, the loads and the fuel consumed by the truck.I want to use the worked hours and loads as the inputs(Two Inputs) and the fuel consumed as the target(output data). my dataset is consist of 80 records and i want to divide it into two sub datasets:training and testing. 65% of the total data for training and the remaining 35% for testing so as to validate accuracy of the model.The number of neuron in the hidden layer should be set to 1 initially and the hyperbolic tangent sigmiod function should be used as the transfer function.The data should be run to be run to give an output with the calculated error between the target and the output(predicted). The error should then be back propagated through the network for adjustments.The process should be repeated by varing the neuron number till the error is minimal. The model should then be simulated to ascertain the validity by using the testing data.The optimim number of neuron that gives the best results should be selected based on the correlation coefficient and the mean absolute percentage error for the testing data.The architecture of the BPNN should be developed.Find the data to be used in is below
WORKED HOURS(Hrs) : 299.9 372.3 401.9 429.5 358.3 79.6 310.3 176.4 221.05 187.6 433.6 243 480.5 203.5 426.8 456.6 473.2 350.4 474 161.4 390.8 471.7 266.3 496.6 423.7 320.9 18.1 504.5 531.5 514.8 519.3 513.2 49 206.3 416.8 290.9 368.9 484.6 209 225.6 439.6 461.9 352.3 455.3 475.6 445.1 346.5 469.7 288.7 306.3 170.6 391.7 147.4 392 343.6 223.6 67.1 361.6 279.5 361.3 451.7 76.6 344.8 352.3 347.7 450.6 434.7 365.9 327.6 451.5 433 361.9 180.3 466.5 415.6 461.7 437.1 478.3 325.2 499.7
LOADS (BCM) 1129 972 703 1199 1438 1144 1092 1526 772 1018 1415 1026 424 931 489 1044 135 181 1488 1038 1112 1191 149 1607 830 944 595 744 1126 1386 922 1217 1471 1124 1490 1436 1095 1266 1245 1384 1183 1299 940 1215 1098 1125 929 1307 1253 1407 1471 836 994 1138 1576 1066 1138 1106 1170 1283 942 1420 663 1167 1118 1254 1214 1187 1387 1311 1108 1336 1082 1435 1326 1188 1527 1036 1172 1517
FUEL CONSUMED (l/bcm)
40,733.00 46,375.00 50,797.00 49,706.00 45,092.00 34,543.87 36,687.62 44,899.04 37,023.81 40,458.24 34,383.52 25,914.49 34,300.42 46,685.87 54,122.83 47,242.32 46,296.99 61,800.78 58,181.10 40,579.12 56,890.57 60,295.81 61,843.33 63,156.89 60,320.82 56,141.04 58,063.53 46,379.00 44,011.00 44,191.00 42,089.00 45,682.00 50,666.00 45,643.00 46,568.00 49,162.00 47,871.63 45,012.70 40,345.86 43,233.45 41,408.20 52,169.66 53,161.51 46,553.05 52,353.93 49,019.43 42,015.50 35,327.23 35,025.50 28,207.78 39,435.27 42,200.30 26,677.27 45,977.46 27,870.61 22,335.77 41,153.24 36,260.02 40,901.63 42,274.00 41,501.00 39,189.00 36,349.00 38,545.00 38,625.00 32,907.00 30,153.00 29,142.58 26,813.88 34,488.11 40,575.67 43,439.41 45,374.04 47,730.75 58,335.01 48,531.21 45,510.96 51,891.60 42,902.32 32,819.21 modeling, machine learning, neural network, fuel vs loads time MATLAB Answers — New Questions
