Search

• Variational Quantum Circuits (VQCs) play a key role in several applications. • VQCs are parameterized quantum circuits trained by using classical optimizers. • The paper proposes to apply memetic algorithms to train VQCs. • The designed memetic algorithm outperforms the state-of-the-art classical optimizers. Among the ready-to-implement quantum algorithms, Variational Quantum Circuits (VQCs) play a key role in several applications, including machine learning. Their strength lies in the use of a parameterized quantum circuit that is trained by means of an optimization algorithm run on a classical computer. In such a scenario, there is a strong need to design appropriate classical optimization schemes that deal efficiently with VQCs and pave the way for quantum advantage in machine learning. Among possible optimization schemes, those based on evolutionary computation are finding increasing interest, given the unconventional and nonanalytical nature of the problem to be solved. This paper proposes to apply memetic algorithms to train VQCs used as quantum classifiers and shows the benefits of exploiting this evolutionary optimization technique through a comparative experimental session. [ABSTRACT FROM AUTHOR]

• Teaching-learning-based optimization (TLBO) is one of population-based heuristic stochastic swarm intelligent algorithm. • In this paper, a comprehensive survey on the recent advances in TLBO is presented. • Literature survey reveals some interesting challenges and future research directions. Teaching-learning-based optimization (TLBO) algorithm which imitates the teaching–learning process in a classroom, is one of population-based heuristic stochastic swarm intelligent algorithms. TLBO executes through similar iterative evolution processes as utilized by a standard evolutionary algorithm. Unlike traditional evolutionary algorithms and swarm intelligent algorithms, the iterative computation process of teaching–learning-based optimization is divided into two phases and each phase executes iterative learning operation. In this paper, we present a comprehensive survey on the recent advances in TLBO. A review of the current literature reveals intriguing challenges and suggests potential future research directions. [ABSTRACT FROM AUTHOR]

Many real-world problems can be abstracted as multimodal global optimization, which is one of the main challenges for optimization algorithms due to its complexity. The fireworks algorithm (FWA) is a swarm intelligence optimization algorithm that has been widely studied and applied by virtue of the synergistic property among fireworks. Current FWA variants have poor exploitation capability to handle some locally complex multimodal functions, which greatly limits the application of the FWA to practical problems. To solve the above problems, in this paper, we propose the multi-guiding spark fireworks algorithm (MGFWA) to solve multimodal functions by enhancing FWA exploitation capabilities. Three different strategies which are boosted guiding vector, multi-guiding sparks, and population-based random mapping are designed to boost the guiding vector, enrich the guiding spark diversity, and fix the mapping function, separately. The validity and parameter setting of MGFWA are theoretically analyzed. Experimentally, the results on the CEC2013 and CEC2017 single objective optimization benchmarks illustrate the remarkable performance of the MGFWA compared to other typical optimization algorithms and FWA variants. Moreover, the ablation study shows each of the three parts plays an important role in the algorithm and the efficiency experiment shows the MGFWA can improve the efficiency of guiding sparks by 10%. We believe that the MGFWA can be considered the SOTA variant of the FWA. • Utilizing boosted guiding vector, multi-guiding sparks, and population-based random mapping to enhance the exploitation capability of the fireworks algorithm. • State-of-the-art performance among fireworks algorithm variants. • Superior performance on multimodal global optimization among several typical swarm intelligence optimization algorithms and evolutionary algorithms. • Extensive experiments (about 60 test functions) and theoretical analyses to demonstrate the effectiveness. [ABSTRACT FROM AUTHOR]

Feature selection algorithms based on evolutionary computation have continued to emerge, and most of them have achieved outstanding results. However, there are two drawbacks when facing high-dimensional datasets: firstly, it is difficult to reduce features effectively, and secondly, the "curse of dimensionality". To alleviate those problems, we take the initial population generation as an entry point and propose a variant initial population generator, which can improve diversity and initialize populations randomly throughout the solution space. However, during the experimental process, it was found that the improved diversity would cause the algorithm to converge too fast and thus lead to premature. Therefore, we introduced multi-population techniques to balance diversity and convergence speed, and finally formed the MPF-FS framework. To prove the effectiveness of this framework, two feature selection algorithms, multi-population multi-objective artificial bee colony algorithm and multi-population non-dominated sorting genetic algorithm II, are implemented based on this framework. Nine well-known public datasets were used in this study, and the results reveal that the two proposed multi-population methods on high-dimensional datasets can reduce more features without reducing (or even improving) classification accuracy, which outperforms the corresponding single-population algorithms. Further compared to the state-of-the-art methods, our method still shows promising results. [ABSTRACT FROM AUTHOR]