Improving flood forecasting through feature selection by a genetic algorithm – experiments based on real data from an Amazon rainforest river

A Novel Physics‐Aware Machine Learning‐Based Dynamic Error Correction Model for Improving Streamflow Forecast Accuracy

“…The time series plot of the error and observed discharge indicated the error sequence follows a definite pattern for the observed discharge (Figures 7g and 7h). Thus, error sequences are certainly demonstrating non‐Gaussian and time‐dependent behavior, and there exists a possibility of error modeling to improve the overall model performance (Cho & Kim, 2022; Farchi et al., 2021; Y. Li et al., 2014; Lipponen et al., 2013; Liu et al., 2015; McInerney et al., 2020; Nester et al., 2012; Shen, Ruijsch, et al., 2022; Shen, Wang, et al., 2022; Tajiki et al., 2020; Tian et al., 2014).…”

“…where e is the error in cumecs, 𝐴𝐴 𝐴𝐴obs is the observed discharge in cumecs and 𝐴𝐴 𝑄𝑄 sim is the mean of the streamflow simulation from prediction interval in cumecs. To assess, if error sequences are normally distributed, a normality test (Anderson Darling test and Kolmogorov-Smirnov test) was performed (Bogner & Pappenberger, 2011;Latt & Wittenberg, 2014;Moeeni & Bonakdari, 2017;Vieira et al, 2021). In addition, several quantitative performance indicators, such as probability density function (PDF) and cumulative distribution function (CDF) of the error sequence were plotted to aid visualization and detect skewness.…”

A Physics‐Aware Machine Learning‐Based Framework for Minimizing Prediction Uncertainty of Hydrological Models

“…The Anderson Darling test and Kolmogorov‐Smirnov test resulted in the non‐acceptance of the null hypothesis (error follows a Gaussian distribution), signifying the error to be non‐Gaussian at 5% significance level (Bogner & Pappenberger, 2011; Latt & Wittenberg, 2014; Moeeni & Bonakdari, 2017; Vieira et al., 2021). The autocorrelation plots of the error reveal the error at a time step “ t ” to have a significant correlation of up to 5 and 15 preceding days in Beas (Figure 7a) and Sunkoshi river basin (Figure 7b) respectively, signifying the error not to be a random white noise.…”

“…The error resulting from the ensemble mean of the streamflow PI is examined for its Gaussian behavior through the normality test like the Anderson Darling test and the Kolmogorov‐Smirnov test (Bogner & Pappenberger, 2011; Latt & Wittenberg, 2014; Moeeni & Bonakdari, 2017; Vieira et al., 2021). Additionally, various graphical performance indicators including the probability density function (PDF) and cumulative distribution function (CDF) of the error were evaluated to understand the Gaussian behavior and skewness involved.…”

A Feature Selection Method by using Chaotic Cuckoo Search Optimization Algorithm with Elitist Preservation and Uniform Mutation for Data Classification

“…ese approaches include genetic algorithms [8][9][10], particle swarm algorithms [11][12][13][14], artificial bee colony algorithms [15], grey wolf optimization algorithms [16], brainstorming algorithms [17], firefly algorithm [18], bat algorithms [19], and butterfly algorithms [20]. Nature-inspired algorithms are committed to the study of exploration and exploitation capabilities.…”