An Optimized Hybrid Forecasting Model and Its Application to Air Pollution Concentration

Abstract

Previous literature suggested that intuitionistic fuzzy inference systems (IFISs) can offer a good forecasting model and intimately linked to the notion of uncertain parameters. However, their performance can be severely degraded by the presence of missing data and less regulated local optima. This study proposes a hybrid IFIS model by assimilating the probabilistic principal component analysis (PPCA) to enhance preprocessing data and particle swarm optimization (PSO) algorithm to optimize the performance of the forecasting model. The main purpose of the PPCA is to diminish outliers affected by defective values and missing values within experimental data. The PSO optimization algorithm is used to tune the parameters of IFIS and thus elevate the prediction performance of the IFIS. Extensive experimental data on meteorological parameters that are recognized as driving factors of tropospheric pollution were employed to study the benefits of the proposed hybrid model. Comparable three error measures are presented to check the performance of the proposed model against the other models. The error analysis result clearly highlights that the proposed hybrid model is performed better compared to the other IFIS-based models and the well-known existing models.