Development of effective and efficient rainfall‐runoff models using integration of deterministic, real‐coded genetic algorithms and artificial neural network techniques

Abstract

[1] Many researchers have reported about the problems in modeling low-magnitude flows while developing artificial neural network (ANN) rainfall-runoff models trained using popular back propagation (BP) method and have suggested the use of alternative training methods. This paper presents the results of a new approach employing real-coded genetic algorithms (GAs) to train ANN rainfall-runoff models, which are able to overcome such problems. The paper also presents a new class of models termed gray box models that integrate deterministic and ANN techniques for hydrologic modeling. Daily rainfall and streamflow data from the Kentucky River watershed were employed to test the new approach. Many standard statistical measures were employed to assess and compare various models investigated. The results obtained in this study demonstrate that ANN rainfall-runoff models trained using real-coded GA are able to predict daily flow more accurately than the ANN rainfall-runoff models trained using BP method. The proposed approach of training ANN models using real-coded GA can significantly improve the estimation accuracy of the low-magnitude flows. It was found that the gray box models that are capable of exploiting the advantages of both deterministic and ANN techniques perform better than the purely black box type ANN rainfall-runoff models. A partitioning analysis of results is needed to evaluate the performance of various models in terms of their efficiency in modeling and effectiveness in accurately predicting varying magnitude flows (low, medium, and high flows).INDEX TERMS: 1860 Hydrology: Runoff and streamflow; 1848 Hydrology: Networks; 1899 Hydrology: General or miscellaneous; KEYWORDS: rainfall-runoff modeling, artificial neural networks, genetic algorithms, optimization, deterministic techniques, conceptual rainfall-runoff models Citation: Jain, A., and S. Srinivasulu (2004), Development of effective and efficient rainfall-runoff models using integration of deterministic, real-coded genetic algorithms and artificial neural network techniques, Water Resour. Res., 40, W04302,

Machine Learning Models Coupled with Variational Mode Decomposition: A New Approach for Modeling Daily Rainfall-Runoff

“…Table 1 summarizes the quantitative performance indices employed in this study. Details on the indices can be found in Jain and Srinivasulu [31] for the effectiveness indices and Dawson et al [77] for efficiency indices, respectively. CE, r 2 , IOA, and PI are dimensionless indices.…”

“…R4MS4E is a good efficiency index for high and peak flows and has the same unit as original data [77,78]. AARE and TS give appropriate weights for low, moderate, and high flows, and it has been reported that they can provide better performance evaluation [31]. …”

“…First, to improve the model performance, the MLMs have been combined with statistical methods, including phase-space reconstruction [22,23], principal component analysis [24,25], fuzzy c-means clustering [7,22], k-means clustering [26,27], self-organizing map (SOM) [28,29] and bootstrap [30]. Second, the MLMs have been coupled with evolutionary optimization algorithms, including genetic algorithm (GA) [31,32], particle swarm optimization (PSO) [11,33], artificial bee colony [34], bat algorithm [35], and firefly algorithm [36]. The addressed algorithms were very helpful for efficient model learning and optimal parameter searching.…”

Discussion on the Choice of Decomposition Level for Wavelet Based Hydrological Time Series Modeling

“…R is an index commonly used to describe the correlation of two series. AARE can measure the error between the observation and simulated data [34]. In general, high NSE and R values and small AARE values indicate good modeling performance.…”

Improving partial mutual information-based input variable selection by consideration of boundary issues associated with bandwidth estimation

“…average daily effective 846 rainfall: P(t), P(t-1) and average daily runoff: Q(t-1), Q(t-2)) (Galelli et al, 2014; Jain and 847 Srinivasulu, 2004). For this case study, the effectiveness of PMI IVS is investigated by 848 introducing another 17 redundant or irrelevant candidate inputs, as shown in Table 6.…”