There are many machine learning approaches available and commonly used today, however, the extreme learning machine is appraised as one of the fastest and, additionally, relatively efficient models. Its main benefit i...
详细信息
There are many machine learning approaches available and commonly used today, however, the extreme learning machine is appraised as one of the fastest and, additionally, relatively efficient models. Its main benefit is that it is very fast, which makes it suitable for integration within products that require models taking rapid decisions. Nevertheless, despite their large potential, they have not yet been exploited enough, according to the recent literature. Extreme learning machines still face several challenges that need to be addressed. The most significant downside is that the performance of the model heavily depends on the allocated weights and biases within the hidden layer. Finding its appropriate values for practical tasks represents an NP-hard continuous optimization challenge. Research proposed in this study focuses on determining optimal or near optimal weights and biases in the hidden layer for specific tasks. To address this task, a multi-swarm hybrid optimization approach has been proposed, based on three swarm intelligence meta-heuristics, namely the artificial bee colony, the firefly algorithm and the sine-cosine algorithm. The proposed method has been thoroughly validated on seven well-known classification benchmark datasets, and obtained results are compared to other already existing similar cutting-edge approaches from the recent literature. The simulation results point out that the suggested multi-swarm technique is capable to obtain better generalization performance than the rest of the approaches included in the comparative analysis in terms of accuracy, precision, recall, and f1-score indicators. Moreover, to prove that combining two algorithms is not as effective as joining three approaches, additional hybrids generated by pairing, each, two methods employed in the proposed multi-swarm approach, were also implemented and validated against four challenging datasets. The findings from these experiments also prove superior performance of the pro
multi-Objective Optimization Problems (MOPS) present several challenges. In particular, when the number of objectives is greater than three, they are actually called Many-Objective Optimization Problems (MaOPs). To ov...
详细信息
ISBN:
(纸本)9781479978595
multi-Objective Optimization Problems (MOPS) present several challenges. In particular, when the number of objectives is greater than three, they are actually called Many-Objective Optimization Problems (MaOPs). To overcome this limitation, researches are investigating multi-swarm approaches. multi-swarm is a very interesting approach that allows the decomposition of different aspects of the problem and each swarm could specialize on a dedicate part of the problem. This study explores this idea to create a novel multi-swarm algorithm, called A-multi, which tackles the main challenge of MaOPs: to convergence towards the true Pareto front and to diversify the obtained solutions covering the entire Pareto front. A-multi project is based on different swarms use different archiving methods, ones specialized on diversity and others specialized on convergence. The algorithm is evaluated with several MaOPs in terms of both convergence and diversity and the results shows the validity of archive based multi-swarm approach.
Dynamic optimization problems (DOPs) are optimization problems that change over time, and most investigations in this area focus on tracking the moving optimum efficiently. However, continuously tracking a moving opti...
详细信息
ISBN:
(纸本)9783319557922;9783319557915
Dynamic optimization problems (DOPs) are optimization problems that change over time, and most investigations in this area focus on tracking the moving optimum efficiently. However, continuously tracking a moving optimum is not practical in many real-world problems because changing solutions frequently is not possible or very costly. Recently, another practical way to tackle DOPs has been suggested: robust optimization over time (ROOT). In ROOT, the main goal is to find solutions that can remain acceptable over an extended period of time. In this paper, a new multi-swarm PSO algorithm is proposed in which different swarms track peaks and gather information about their behavior. This information is then used to make decisions about the next robust solution. The main goal of the proposed algorithm is to maximize the average number of environments during which the selected solutions' quality remains acceptable. The experimental results show that our proposed algorithm can perform significantly better than existing work in this aspect.
暂无评论