Evolutionary Computation最新文献

Heuristic optimization methods such as Particle Swarm Optimization depend on their parameters to achieve optimal performance on a given class of problems. Some modifications of heuristic algorithms aim at adapting those parameters during the optimization process. We present a novel approach to design such adaptation strategies using continuous fuzzy feedback control. Fuzzy feedback provides a simple interface where probes are sampled in the optimization process and parameters are fed back to the optimizer. The probes are turned into parameters by a fuzzy process optimized beforehand to maximize performance on a training benchmark. Utilizing this framework, we systematically established 127 different Fuzzy Particle Swarm Optimization algorithms featuring a maximum of 7 parameters under fuzzy control. These newly devised algorithms exhibit superior performance compared to both traditional PSO and some of its best parameter control variants. The performance is reported in the single-objective bound-constrained numerical optimization competition of CEC 2020. Additionally, two specific controls, highlighted for their efficacy and dependability, demonstrated commendable performance in real-world scenarios from CEC 2011.

Recently, computationally intensive multiobjective optimization problems have been efficiently solved by surrogate-assisted multiobjective evolutionary algorithms. However, most of those algorithms can only handle no more than 200 decision variables. As the number of decision variables increases further, unreliable surrogate models will result in a dramatic deterioration of their performance, which makes large-scale expensive multiobjective optimization challenging. To address this challenge, we develop a large-scale multiobjective evolutionary algorithm guided by low-dimensional surrogate models of scalarization functions. The proposed algorithm (termed LDS-AF) reduces the dimension of the original decision space based on principal component analysis, and then directly approximates the scalarization functions in a decompositionbased multiobjective evolutionary algorithm. With the help of a two-stage modeling strategy and convergence control strategy, LDS-AF can keep a good balance between convergence and diversity, and achieve a promising performance without being trapped in a local optimum prematurely. The experimental results on a set of test instances have demonstrated its superiority over eight state-of-the-art algorithms on multiobjective optimization problems with up to 1000 decision variables using only 500 real function evaluations.

Evolution-based neural architecture search methods have shown promising results, but they require high computational resources because these methods involve training each candidate architecture from scratch and then evaluating its fitness, which results in long search time. Covariance Matrix Adaptation Evolution Strategy (CMA-ES) has shown promising results in tuning hyperparameters of neural networks but has not been used for neural architecture search. In this work, we propose a framework called CMANAS which applies the faster convergence property of CMA-ES to the deep neural architecture search problem. Instead of training each individual architecture seperately, we used the accuracy of a trained one shot model (OSM) on the validation data as a prediction of the fitness of the architecture, resulting in reduced search time. We also used an architecture-fitness table (AF table) for keeping a record of the already evaluated architecture, thus further reducing the search time. The architectures are modeled using a normal distribution, which is updated using CMA-ES based on the fitness of the sampled population. Experimentally, CMANAS achieves better results than previous evolution-based methods while reducing the search time significantly. The effectiveness of CMANAS is shown on two different search spaces using four datasets: CIFAR-10, CIFAR-100, ImageNet, and ImageNet16-120. All the results show that CMANAS is a viable alternative to previous evolution-based methods and extends the application of CMA-ES to the deep neural architecture search field.

We contribute to the efficient approximation of the Pareto-set for the classical NP-hard multiobjective minimum spanning tree problem (moMST) adopting evolutionary computation. More precisely, by building upon preliminary work, we analyze the neighborhood structure of Pareto-optimal spanning trees and design several highly biased sub-graph-based mutation operators founded on the gained insights. In a nutshell, these operators replace (un)connected sub-trees of candidate solutions with locally optimal sub-trees. The latter (biased) step is realized by applying Kruskal's single-objective MST algorithm to a weighted sum scalarization of a sub-graph. We prove runtime complexity results for the introduced operators and investigate the desirable Pareto-beneficial property. This property states that mutants cannot be dominated by their parent. Moreover, we perform an extensive experimental benchmark study to showcase the operator's practical suitability. Our results confirm that the sub-graph-based operators beat baseline algorithms from the literature even with severely restricted computational budget in terms of function evaluations on four different classes of complete graphs with different shapes of the Pareto-front.

Exposing an evolutionary algorithm that is used to evolve robot controllers to variable conditions is necessary to obtain solutions which are robust and can cross the reality gap. However, we do not yet have methods for analyzing and understanding the impact of the varying morphological conditions which impact the evolutionary process, and therefore for choosing suitable variation ranges. By morphological conditions, we refer to the starting state of the robot, and to variations in its sensor readings during operation due to noise. In this paper, we introduce a method that permits us to measure the impact of these morphological variations and we analyze the relation between the amplitude of variations, the modality with which they are introduced, and the performance and robustness of evolving agents. Our results demonstrate that (i) the evolutionary algorithm can tolerate morphological variations which have a very high impact, (ii) variations affecting the actions of the agent are tolerated much better than variations affecting the initial state of the agent or of the environment, and (iii) improving the accuracy of the fitness measure through multiple evaluations is not always useful. Moreover, our results show that morphological variations permit generating solutions which perform better both in varying and non-varying conditions.

In this paper, we compare Bayesian Optimization, Differential Evolution, and an Evolution Strategy employed as a gait-learning algorithm in modular robots. The motivational scenario is the joint evolution of morphologies and controllers, where "newborn" robots also undergo a learning process to optimize their inherited controllers (without changing their bodies). This context raises the question: How do gait-learning algorithms compare when applied to various morphologies that are not known in advance (and thus need to be treated as without priors)? To answer this question, we use a test suite of twenty different robot morphologies to evaluate our gait-learners and compare their efficiency, efficacy, and sensitivity to morphological differences. The results indicate that Bayesian Optimization and Differential Evolution deliver the same solution quality (walking speed for the robot) with fewer evaluations than the Evolution Strategy. Furthermore, the Evolution Strategy is more sensitive for morphological differences (its efficacy varies more between different morphologies) and is more subject to luck (repeated runs on the same morphology show greater variance in the outcomes).

Premature convergence is a thorny problem for particle swarm optimization (PSO) algorithms, especially on multimodal problems, where maintaining swarm diversity is crucial. However, most enhancement strategies for PSO, including the existing diversity-guided strategies, have not fully addressed this issue. This paper proposes the virtual position guided (VPG) strategy for PSO algorithms. The VPG strategy calculates diversity values for two different populations and establishes a diversity baseline. It then dynamically guides the algorithm to conduct different search behaviors, through three phases - divergence, normal, and acceleration - in each iteration, based on the relationships among these diversity values and the baseline. Collectively, these phases orchestrate different schemes to balance exploration and exploitation, collaboratively steering the algorithm away from local optima and towards enhanced solution quality. The introduction of 'virtual position' caters to the strategy's adaptability across various PSO algorithms, ensuring the generality and effectiveness of the proposed VPG strategy. With a single hyperparameter and a recommended usual setup, VPG is easy to implement. The experimental results demonstrate that the VPG strategy is superior to several canonical and the state-of-the-art strategies for diversity guidance, and is effective in improving the search performance of most PSO algorithms on multimodal problems of various dimensionalities.

Gathering sufficient instance data to either train algorithm-selection models or understand algorithm footprints within an instance space can be challenging. We propose an approach to generating synthetic instances that are tailored to perform well with respect to a target algorithm belonging to a predefined portfolio but are also diverse with respect to their features. Our approach uses a novelty search algorithm with a linearly weighted fitness function that balances novelty and performance to generate a large set of diverse and discriminatory instances in a single run of the algorithm. We consider two definitions of novelty: (1) with respect to discriminatory performance within a portfolio of solvers; (2) with respect to the features of the evolved instances. We evaluate the proposed method with respect to its ability to generate diverse and discriminatory instances in two domains (knapsack and bin-packing), comparing to another well-known quality diversity method, Multi-dimensional Archive of Phenotypic Elites (MAP-Elites) and an evolutionary algorithm that only evolves for discriminatory behaviour. The results demonstrate that the novelty search method outperforms its competitors in terms of coverage of the space and its ability to generate instances that are diverse regarding the relative size of the "performance gap" between the target solver and the remaining solvers in the portfolio. Moreover, for the Knapsack domain, we also show that we are able to generate novel instances in regions of an instance space not covered by existing benchmarks using a portfolio of state-of-the-art solvers. Finally, we demonstrate that the method is robust to different portfolios of solvers (stochastic approaches, deterministic heuristics and state-of-the-art methods), thereby providing further evidence of its generality.

Evolutionary Computation (EC) often throws away learned knowledge as it is reset for each new problem addressed. Conversely, humans can learn from small-scale problems, retain this knowledge (plus functionality) and then successfully reuse them in larger-scale and/or related problems. Linking solutions to problems together has been achieved through layered learning, where an experimenter sets a series of simpler related problems to solve a more complex task. Recent works on Learning Classifier Systems (LCSs) has shown that knowledge reuse through the adoption of Code Fragments, GP-like tree-based programs, is plausible. However, random reuse is inefficient. Thus, the research question is how LCS can adopt a layered-learning framework, such that increasingly complex problems can be solved efficiently? An LCS (named XCSCF*) has been developed to include the required base axioms necessary for learning, refined methods for transfer learning and learning recast as a decomposition into a series of subordinate problems. These subordinate problems can be set as a curriculum by a teacher, but this does not mean that an agent can learn from it. Especially if it only extracts over-fitted knowledge of each problem rather than the underlying scalable patterns and functions. Results show that from a conventional tabula rasa, with only a vague notion of what subordinate problems might be relevant, XCSCF* captures the general logic behind the tested domains and therefore can solve any n-bit Multiplexer, n-bit Carry-one, n-bit Majority-on, and n-bit Even-parity problems. This work demonstrates a step towards continual learning as learned knowledge is effectively reused in subsequent problems.

We study the (1:s+1) success rule for controlling the population size of the (1,λ)- EA. It was shown by Hevia Fajardo and Sudholt that this parameter control mechanism can run into problems for large s if the fitness landscape is too easy. They conjectured that this problem is worst for the ONEMAX benchmark, since in some well-established sense ONEMAX is known to be the easiest fitness landscape. In this paper we disprove this conjecture. We show that there exist s and ɛ such that the self-adjusting (1,λ)-EA with the (1:s+1)-rule optimizes ONEMAX efficiently when started with ɛn zero-bits, but does not find the optimum in polynomial time on DYNAMIC BINVAL. Hence, we show that there are landscapes where the problem of the (1:s+1)-rule for controlling the population size of the (1,λ)-EA is more severe than for ONEMAX. The key insight is that, while ONEMAX is the easiest function for decreasing the distance to the optimum, it is not the easiest fitness landscape with respect to finding fitness-improving steps.