Search

Seepage prediction is a vital part of the dam safety monitoring system. Traditional statistical models ignore the nonlinear characteristics of the measured variables, resulting in poor accuracy and stability. In this study, an optimized neural network model is constructed to predict the seepage extent of hydropower station dams. An improved gradient-based optimizer (IGBO) is proposed to increase the accuracy and reliability of extreme learning machine (ELM) model predictions. The IGBO introduces an initialization method with elite opposition-based learning to improve population diversity. A crossover operator and a nonlinear parameter are used in the IGBO to enhance the ability of local search and the probability of avoiding local optima. The performance of the IGBO-optimized ELM network (IGBO-ELM) was evaluated on 12 datasets. In addition, the comparison experimental results with actual monitoring data of concrete dams show that IGBO-ELM has strong generalization performance and accuracy among the other four optimization models. [ABSTRACT FROM AUTHOR]

The hybrid models of granular computing and support vector machine are a kind of new machine learning algorithms based on granular computing and statistical learning theory. These hybrid models can effectively use the advantage of each algorithm, so that their performance are better than a single method. In view of their excellent learning performance, the hybrid models of granular computing and support vector machine have become one of the focus at home and abroad. In this paper, the research on the hybrid models are reviewed, which include fuzzy support vector machine, rough support vector machine, quotient space support vector machine, rough fuzzy support vector machine and fuzzy rough support vector machine. Firstly, we briefly introduce the typical granular computing models and the basic theory of support vector machines. Secondly, we describe the latest progress of these hybrid models in recent years. Finally, we point out the research and development prospects of the hybrid algorithms. [ABSTRACT FROM AUTHOR]