Research Article

Hybrid Machine Learning Model for Electricity Consumption Prediction Using Random Forest and Artificial Neural Networks

Table 1

Descriptive statistics of variables for the electricity consumption prediction.

Group of variablesVariable descriptionVariable codeDescriptive statistics

GeospatialElectrical substation (0–2 = south area, 1–3 3–5 = north area, 1–3 6–8 = northeast area, 1–3 9–11 = central area 1–3)Electrical_SubstationCategorical, 0 = 8.35%, 1 = 8.35, 2 = 8.35%, 3 = 12.52%, 4 = 8.35%, 5 = 8.35%, 6 = 8.40%, 7 = 8.35%, 8 = 8.22%, 9 = 8.35%, 10 = 4.07%, 11 = 8.35%
Business type (0–1 = small, large residential houses, 2–4 = small, medium, large business, 5 = specific activities, 6 = government, 7 = agriculture)Usage_TypeCategorical, 0 = 13.02%, 1 = 12.91%, 2 = 12.05%, 3 = 12.20%, 4 = 12.00%, 5 = 12.00%, 6 = 12.78%, 7 = 13.04%
Time of use (0 = peakday, 1 = semipeak, 2 = offpeak)TOUCategorical, 0 = 12.50%, 1 = 66.67%, 2 = 20.83%
Thailand season period (0 = summer, 1 = rainy, 2 = winter)SeasonCategorical, 0 = 24.50%, 1 = 42.94%, 2 = 32.56%
GeographicalTotal number of populationsPopulation_NNumerical, min = 2810387.00, max = 7871210.00, mean = 4965500.09, st. dev = 1641653.19
Total surface areaAreaNumerical, min = 22423.00, max = 72806.35, mean = 42012.25, st. dev = 16950.41
Ratio of people and area per sq. kmPopulation_RatioNumerical, min = 445.00, max = 1860.00, mean = 1003.87, st. dev = 369.42
ClimaticMean station pressureMean_Station_PressureNumerical, min = 975.23, max = 1010.86, mean = 999.44, st. dev = 9.18
Mean msl pressureMean_MSL_PressureNumerical, min = 1004.46, max = 1014.20, mean = 1008.85, st. dev = 2.50
Mean maximum temperatureMean_Maximum_TemperatureNumerical, min = 27.92, max = 36.62, mean = 32.15, st. dev = 1.71
Mean minimum temperatureMean_Minimum_TemperatureNumerical, min = 14.00, max = 25.36, mean = 22.24, st. dev = 2.56
Mean drybulb temperatureMean_Drybulb_TemperatureNumerical, min = 20.46, max = 29.87, mean = 26.59, st. dev = 1.84
Mean relative humidityMean_Relative_HumidityNumerical, min = 57.00, max = 85.73, mean = 75.56, st. dev = 6.36
Total rainfallTotal_RainfallNumerical, min = 11.00, max = 4841.00, mean = 1264.52, st. dev = 1096.77
IndustrialTotal number of industrial laborsIndustrial_LaborNumerical, min = 1471.00, max = 21368.00, mean = 7087.11, st. dev = 6573.51
Total number of industrial plantsIndustrial_PlantNumerical, min = 101.00, max = 727.00, mean = 263.75, st. dev = 167.60
HouseholdTotal number of agriculturistsAgriculturistsNumerical, min = 197885.00, max = 1412997.00, mean = 628382.35, st. dev = 375015.78
Average expenditure per householdExpenditureNumerical, min = 63795.00, max = 168442.00, mean = 112928.45, st. dev = 28137.06
Average income per householdIncomeNumerical, min = 81571.00, max = 206231.00, mean = 143155.35, st. dev = 34841.15
Total number of householdsHouseholdNumerical, min = 861.00, max = 1984.00, mean = 1479.48, st. dev = 318.94
Average liabilities per householdLiabilitiesNumerical, min = 383606.00, max = 1536859.00, mean = 1005455.20, st. dev = 316864.28
Target variableElectricity consumptionDemand_KWNumerical, min = 0.10, max = 8731298.26, mean = 167920.67, st. dev = 358252.18