A Novel Nonparametric Regression Ensemble for Rainfall Forecasting Using Particle Swarm Optimization Technique Coupled with Artificial Neural Network

Abstract

Abstract. In this study, we propose a novel nonparametric regression (NR) ensemble rainfall forecasting model integrating generalized particle swarm optimization (PSO) with artificial neural network (ANN). First of all, the PSO algorithm is used to evolve neural network architecture and connection weights. The evolved neural network architecture and connection weights are input into a new neural network.The new neural network is trained using back-propagation (BP) algorithm, generating different individual neural network. Then, the principal component analysis (PCA) technology is adopted to extract ensemble members. Finally, the NR is used for nonlinear ensemble model. Empirical results obtained reveal that the prediction by using the NR ensemble model is generally better than those obtained using other models presented in this study in terms of the same evaluation measurements. For illustration and testing reveal that the NR ensemble model proposed can be used as an alternative forecasting tool for a Meteorological application in achieving greater forecasting accuracy and improving prediction quality further.

Support Vector Regression Based on Particle Swarm Optimization for Rainfall Forecasting

“…In order to measure effectiveness of the proposed method, we compare results of BNN ensemble. Four types of errors, such as, the mean absolute percentage error (MAPE), the root mean squares error (RMSE), the mean absolute error (MAE) and Pearson Relative Coefficient (PR) which have been found in many papers, are also used here, Interested readers are referred to [16] for more details.…”

“…Secondly, the SVR training process is equivalent to solving linearly constrained quadratic programming problems and the SVR embedded solution meaning is unique, optimal and unlikely to generate local minima. Finally, it chooses only the necessary data points to solve the regression function, which results in the sparseness of solution [13], [14], [15].…”

“…The success of the ANN models is attributable to its generalization capability to predict the output of new data after the neural network is trained [1], [2]. Some of these studies, however, showed that ANN had some limitations in learning the patterns because the generalization of single neural network is not unique [3], [4]. In addition, ANN approaches want of a strict theoretical support, effects of applications are strongly depended upon operator's experience.…”

A Novel Nonlinear Combination Model Based on Support Vector Machine for Rainfall Prediction

“…With the development of science and technology, in particular, the intelligent computing technology in the past few decades, many emerging techniques, such as artificial neural network (ANN), have been widely used in the rainfall forecasting and obtained good results [3], [4], [5]. ANN are computerized intelligence systems that simulate the inductive power and behavior of the human brain.…”

A Novel Bayesian Additive Regression Trees Ensemble Model Based on Linear Regression and Nonlinear Regression for Torrential Rain Forecasting

“…Although the climate dynamics method for rainfall forecasting has been making great progress, given the short time scale, the small catchments area, and the massive costs associated with collecting the required meteorological data, it is not a feasible alternative in most cases because it involves many variables which are interconnected in a very complicated way. Even if the weather dynamics model can be the effective establishment, it is difficult to solve because it involves a complex numerical model of computing technology [2], [3].…”