A hybrid computational intelligence method for predicting dew point temperature

Recently, utilization of hybrid models has gained remarkable attention as they take the advantage of specific nature of each technique to enhance the precision and reliability of the predictions. In this research work, a new hybrid approach combined the extreme learning machine (ELM) with wavelet transform (WT) algorithm is proposed to predict daily dew point temperature. To test the validity of the proposed approach, the daily weather data sets for port of Bandar Abass situated in the south costal part of Iran are used. The merit of the proposed ELM-WT method is verified against the ELM, support vector machines and artificial neural network techniques based upon several well-known statistical indicators. The achieved results demonstrate that the hybrid ELM-WT method presents absolute superiority over other powerful techniques applied. It is found that among four considered sets of parameters with 1, 2 and 3 inputs, further accuracy can be achieved using combination of average ambient temperature (T _avg) and relative humidity (R _h). For the best ELM-WT model using T _avg and R _h as inputs, the statistical indicators of mean absolute percentage error, mean absolute bias error, root mean square error and coefficient of determination are 6.1664 %, 0.5495, 0.7621 and 0.9953 °C, respectively. Based on relative percentage error (RPE), for the best ELM-WT model 91 % of the predictions fall within the RPE acceptable range of −10 and +10 %. In a nutshell, the results of this study convincingly advocate that coupling the ELM with WT would be particularly appealing to offer accurate predictions and favorable enhancement in the precision of ELM.