Showing papers presented at "Congress on Evolutionary Computation in 2017"

Single objective real-parameter optimization: Algorithm jSO

Abstract: Solving single objective real-parameter optimization problems, also known as a bound-constrained optimization, is still a challenging task. We can find such problems in engineering optimization, scientific applications, and in other real-world problems. Usually, these problems are very complex and computationally expensive. A new algorithm, called jSO, is presented in this paper. The algorithm is an improved variant of the iL-SHADE algorithm, mainly with a new weighted version of mutation strategy. The experiments were performed on CEC 2017 benchmark functions, which are different from previous competition benchmark functions. A comparison of the proposed jSO algorithm and the L-SHADE algorithm is presented first. From the obtained results we can conclude that jSO performs better in comparison with the L-SHADE algorithm. Next, a comparison of jSO and iL-SHADE is also performed, and jSO obtained better or competitive results. Using the CEC 2017 evaluation method, jSO obtained the best final score among these three algorithms.

Ensemble sinusoidal differential covariance matrix adaptation with Euclidean neighborhood for solving CEC2017 benchmark problems

Abstract: Many Differential Evolution algorithms are introduced in the literature to solve optimization problems with diverse set of characteristics. In this paper, we propose an extension of the previously published paper LSHADE-EpSin that was ranked as the joint winner in the real-parameter single objective optimization competition, CEC 2016. The contribution of this work constitutes two major modifications that have been added to enhance the performance: ensemble of sinusoidal approaches based on performance adaptation and covariance matrix learning for the crossover operator. Two sinusoidal waves have been used to adapt the scaling factor: non-adaptive sinusoidal decreasing adjustment and an adaptive sinusoidal increasing adjustment. Instead of choosing one of the sinusoidal waves randomly, a performance adaptation scheme based on earlier success is used in this work. Moreover, covariance matrix learning with Euclidean neighborhood is used for the crossover operator to establish a suitable coordinate system, and to enhance the capability of LSHADE-EpSin to tackle problems with high correlation between the variables. The proposed algorithm, namely LSHADE-cnEpSin, is tested on the IEEE CEC2017 problems used in the Special Session and Competitions on Single Objective Bound Constrained Real-Parameter Single Objective Optimization. The results statistically affirm the efficiency of the proposed approach to obtain better results compared to other state-of-the-art algorithms.

LSHADE with semi-parameter adaptation hybrid with CMA-ES for solving CEC 2017 benchmark problems

Abstract: To improve the optimization performance of LSHADE algorithm, an alternative adaptation approach for the selection of control parameters is proposed. The proposed algorithm, named LSHADE-SPA, uses a new semi-parameter adaptation approach to effectively adapt the values of the scaling factor of the Differential evolution algorithm. The proposed approach consists of two different settings for two control parameters F and Cr. The benefit of this approach is to prove that the semi-adaptive algorithm is better than pure random algorithm or fully adaptive or self-adaptive algorithm. To enhance the performance of our algorithm, we also introduced a hybridization framework named LSHADE-SPACMA between LSHADE-SPA and a modified version of CMA-ES. The modified version of CMA-ES undergoes the crossover operation to improve the exploration capability of the proposed framework. In LSHADE-SPACMA both algorithms will work simultaneously on the same population, but more populations will be assigned gradually to the better performance algorithm. In order to verify and analyze the performance of both LSHADE-SPA and LSHADE-SPACMA, Numerical experiments on a set of 30 test problems from the CEC2017 benchmark for 10, 30, 50 and 100 dimensions, including a comparison with LSHADE algorithm are executed. Experimental results indicate that in terms of robustness, stability, and quality of the solution obtained, of both LSHADE-SPA and LSHADE-SPACMA are better than LSHADE algorithm, especially as the dimension increases.

Improving the local search capability of Effective Butterfly Optimizer using Covariance Matrix Adapted Retreat Phase

Abstract: Effective Butterfly Optimizer(EBO) is a self-adaptive Butterfly Optimizer which incorporates a crossover operator in Perching and Patrolling to increase the diversity of the population. This paper proposes a new retreat phase called Covariance Matrix Adapted Retreat Phase (CMAR), which uses covariance matrix to generate a new solution and thus improves the local search capability of EBO. This version of EBO is called EBOwithCMAR. We evaluated the performance of EBOwithCMAR on CEC-2017 benchmark problems and compared with the results of winners of a special session of CEC-2016 for bound-constrained problems. The experimental results show that EBOwithCMAR is competitive with the compared algorithms.

An empirical study of multifactorial PSO and multifactorial DE

Abstract: Recently, the notion of Multifactorial Optimization (MFO) has emerged as a promising approach for evolutionary multi-tasking by automatically exploiting the latent synergies between optimization problems, simply through solving them together in an unified representation space [1]. It aims to improve convergence characteristics across multiple optimization problems at once by seamlessly transferring knowledge between them. In [1], the efficacy of MFO has been studied by a specific mode of knowledge transfer in the form of implicit genetic transfer through chromosomal crossover. Here we further explore the generality of MFO when diverse population based search mechanisms are employed. In particular, in this paper, we present the first attempt to conduct MFO with the popular particle swarm optimization and differential evolution search. Two specific multi-tasking paradigms, namely multifactorial particle swarm optimization (MFPSO) and multifactorial differential evolution (MFDE) are proposed. To evaluate the performance of MFPSO and MFDE, comprehensive empirical studies on 9 single objective MFO benchmark problems are provided.

Linearized domain adaptation in evolutionary multitasking

Abstract: Recent analytical studies have revealed that in spite of promising success in problem solving, the performance of evolutionary multitasking deteriorates with decreasing similarity between constitutive tasks. The present day multifactorial evolutionary algorithm (MFEA) is susceptible to negative knowledge transfer between uncorrelated tasks. To alleviate this issue, we propose a linearized domain adaptation (LDA) strategy that transforms the search space of a simple task to the search space similar to its constitutive complex task. This high order representative space resembles high correlation with its constitutive task and provides a platform for efficient knowledge transfer via crossover. The proposed framework, LDA-MFEA is tested on several benchmark problems constituting of tasks with different degrees of similarities and intersecting global optima. Experimental results demonstrate competitive performances against MFEA and shows that our proposition dramatically improves the performance relative to optimizing each task independently.

Parting ways and reallocating resources in evolutionary multitasking

Abstract: Evolutionary multitasking aims to explore implicit synergy among multiple optimization tasks. Through the effect of hitchhiking, evolutionary multitasking is capable of improving the performance of evolutionary algorithms on exploration as well as exploitation. Multifactorial evolutionary algorithm (MFEA) presented an effectual implementation of evolutionary multitasking, which simultaneously seeks the solutions to multiple optimization problems by unifying their search spaces. The MFEA enables information sharing across tasks during evolution. This mechanism can improve the evolutionary efficiency in the early phase; however, it will impair the exploitation and consume extra resources later on, due to the essential difference among the fitness landscapes of optimization problems. This study proposes detecting the occurrence of parting ways, at which the information sharing begins to fail. In addition, we develop the resource allocation mechanism to reallocate the fitness evaluations on different types of offspring by ceasing information sharing when parting ways. Experiments are conducted to evaluate the proposed methods. The experimental results show that applying parting ways detection and resource reallocation for MFEA can achieve better solution quality in most of testing cases, especially when the tasks share low similarity of landscapes.

A unified differential evolution algorithm for constrained optimization problems

Abstract: In this paper, a unified differential evolution algorithm, named UDE, is presented for real parameter constrained optimization problems. The proposed UDE algorithm is inspired from some popular DE variants existing in the literature such as CoDE, JADE, SaDE, and ranking-based mutation operator. The primary feature of UDE lies in unifying the main idea of CoDE, JADE, SaDE, and ranking-based mutation. UDE uses three trial vector generation strategies and two parameter settings. At each generation, UDE divides the current population into two sub-populations. In the top sub-population, UDE employs all the three trial vector generation strategies on each target vector, just like in CoDE. For the bottom sub-population, UDE employs strategy adaptation, in which the trial vector generation strategies are periodically self-adapted by learning from their experiences in generating promising solutions in the top sub-population. Further, UDE utilizes a DE mutation strategy based local search operation. The constraints are handled in UDE using static penalty method. In contrast to most of the DE variants presented in the literature, UDE employs a generational replacement strategy. The proposed UDE algorithm is tested on the 28 benchmark problems provided for the CEC 2017 competition on constrained real parameter optimization. The experimental results demonstrate the efficacy of the presented algorithm in solving constrained real parameter optimization problems.

L-SHADE with competing strategies applied to constrained optimization

Abstract: An enhanced version of L-SHADE algorithm has been proposed recently. The LSHADE44 algorithm uses three different additional strategies for computing a trial point and employed strategies compete in the algorithm. The algorithm was originally developed for bound-constrained optimization. In this paper, the LSHADE44 algorithm was slightly simplified and modified to be able to solve constrained problems. The benchmark set arranged for CEC2017 competition on constrained real-parameter optimization is employed.

Evolutionary many-tasking based on biocoenosis through symbiosis: A framework and benchmark problems

Abstract: Evolutionary multitasking is an emergent topic in evolutionary computation area. Recently, a well-known evolutionary multitasking method, the multi-factorial evolutionary algorithm (MFEA), has been proposed and applied to concurrently solve two or three problems. In MFEA, individuals of different tasks are recombined in a predefined random mating probability. As the number of tasks increases, such recombination of different tasks becomes very frequent, thereby detracting the search from any specific problems and limiting the MFEA's capability to solve many-tasking problems. This study proposes a general framework, called the evolution of biocoenosis through symbiosis (EBS), for evolutionary algorithms to deal with the many-tasking problems. The EBS has two main features: the selection of candidates from concatenate offspring and the adaptive control of information exchange among tasks. The concatenate offspring represent a set of offspring used for all tasks. Moreover, this study presents a test suite of many-tasking problems (MaTPs), modified from the CEC 2014 benchmark problems. The Spearman correlation is adopted to analyze the effect of the shifts of optima on the MaTPs. Experimental results show that the effectiveness of EBS is superior to that of single task optimization and MFEA on the four MaTPs. The results also validate that EBS is capable of exploiting the synergy of fitness landscapes.

A comparative study of constrained multi-objective evolutionary algorithms on constrained multi-objective optimization problems

Abstract: Solving constrained multi-objective optimization problems is a difficult task, it needs to simultaneously optimize multiple conflicting objectives and a number of constraints. This paper first reviews a number of popular constrained multi-objective evolutionary algorithms (CMOEAs) and twenty-three widely used constrained multi-objective optimization problems (CMOPs) (including CF1-10, CTP1-8, BNH, CONSTR, OSY, SRN and TNK problems). Then eight popular CMOEAs with simulated binary crossover (SBX) and differential evolution (DE) operators are selected to test their performance on the twenty-three CMOPs. The eight CMOEAs can be classified into domination-based CMOEAs (including ATM, IDEA, NSGA-II-CDP and SP) and decomposition-based CMOEAs (including CMOEA/D, MOEA/D-CDP, MOEA/D-SR and MOEA/D-IEpsilon). The comprehensive experimental results indicate that IDEA has the best performance in the domination-based CMOEAs and MOEA/D-IEpsilon has the best performance in the decomposition-based CMOEAs. Among the eight CMOEAs, MOEA/D-IEpsilon with both SBX and DE operators has the best performance on the twenty-three test problems.

Adaptive constraint handling and Success History Differential Evolution for CEC 2017 Constrained Real-Parameter Optimization

Abstract: This paper presents Success-History Based Adaptive Differential Evolution Algorithm (SHADE) including Linear population size reduction (L-SHADE), enhanced with adaptive constraint violation handling, applied to the benchmark for CEC 2017 Competition on Constrained Real-Parameter Optimization. The constraint handling method prioritizes the feasible solutions before infeasible, while disregarding the constraint violation values below an adaptive threshold, i.e. adaptive ϵ-constraint handling. The 28 constrained test functions on 10, 30, 50, and 100 dimensions are assessed on the benchmark and the required resulting final fitnesses, constraints violations, and success rates are reported for 25 independent runs of the proposed algorithm under the budget of fixed maximum number of fitness evaluations for 10, 30, 50, and 100 dimensional test functions.

Proactive Particles in Swarm Optimization: A settings-free algorithm for real-parameter single objective optimization problems

Abstract: Particle Swarm Optimization (PSO) is an effective Swarm Intelligence technique for the optimization of non-linear and complex high-dimensional problems. Since PSO's performance is strongly dependent on the choice of its functioning settings, in this work we consider a self-tuning version of PSO, called Proactive Particles in Swarm Optimization (PPSO). PPSO leverages Fuzzy Logic to dynamically determine the best settings for the inertia weight, cognitive factor and social factor. The PPSO algorithm significantly differs from other versions of PSO relying on Fuzzy Logic, because specific settings are assigned to each particle according to its history, instead of being globally assigned to the whole swarm. In such a way, PPSO's particles gain a limited autonomous and proactive intelligence with respect to the reactive agents proposed by PSO. Our results show that PPSO achieves overall good optimization performances on the benchmark functions proposed in the CEC 2017 test suite, with the exception of those based on the Schwefel function, whose fitness landscape seems to mislead the fuzzy reasoning. Moreover, with many benchmark functions, PPSO is characterized by a higher speed of convergence than PSO in the case of high-dimensional problems.

Learning heuristic selection using a Time Delay Neural Network for Open Vehicle Routing

Abstract: A selection hyper-heuristic is a search method that controls a prefixed set of low-level heuristics for solving a given computationally difficult problem. This study investigates a learning-via demonstrations approach generating a selection hyper-heuristic for Open Vehicle Routing Problem (OVRP). As a chosen ‘expert’ hyper-heuristic is run on a small set of training problem instances, data is collected to learn from the expert regarding how to decide which low-level heuristic to select and apply to the solution in hand during the search process. In this study, a Time Delay Neural Network (TDNN) is used to extract hidden patterns within the collected data in the form of a classifier, i.e an ‘apprentice’ hyper-heuristic, which is then used to solve the ‘unseen’ problem instances. Firstly, the parameters of TDNN are tuned using Taguchi orthogonal array as a design of experiments method. Then the influence of extending and enriching the information collected from the expert and fed into TDNN is explored on the behaviour of the generated apprentice hyper-heuristic. The empirical results show that the use of distance between solutions as an additional information collected from the expert generates an apprentice which outperforms the expert algorithm on a benchmark of OVRP instances.

A simple framework for constrained problems with application of L-SHADE44 and IDE

Abstract: A simple framework for cooperation of evolutionary algorithms in the solution of constrained optimization problems is proposed and implemented using two advanced adaptive variants of differential evolution. The new algorithm is applied to the test problems defined for CEC 2017 competition on constrained single objective real-parameter optimization. The new algorithm finds acceptable solution in 75 % of test problems in the all tested dimensions with high efficiency. However, the algorithm is not able to find any feasible solution in the rest 25 % of the test problems. This issue is a field for next research.

A version of IPOP-CMA-ES algorithm with midpoint for CEC 2017 single objective bound constrained problems

Abstract: This paper presents the RB-IPOP-CMA-ES algorithm which is an enhanced version of IPOP-CMA-ES. The algorithm uses midpoint of the population as an approximation of the optimum. The midpoint fitness is also used to introduce a new restart trigger for IPOP. Other IPOP restart triggers and parameters are also corrected. The performance of the proposed approach is evaluated on 30 problems from the CEC 2017 benchmark for 10, 30, 50 and 100 dimensions. The results confirm that RB-IPOP-CMA-ES achieves better results than its version that does not utilize midpoint and is a considerable improvement over a plain IPOP-CMA-ES.

Optimal placement of wind turbines in a windfarm using L-SHADE algorithm

Abstract: Setting of turbines in a windfarm is a complex task as several factors need to be taken into consideration. During recent years, researchers have applied various evolutionary algorithms to windfarm layout problem by converting it to a single objective and at the most two objective optimization problem. The prime factor governing placement of turbines is the wake effect attributed to the loss of kinetic energy by wind after it passes over a turbine. Downstream turbine inside the wake region generates less output power. Optimizing the wake loss helps extract more power out of the wind. The cost of turbine is tactically entwined with generated output to form single objective of cost per unit of output power e.g. cost/kW. This paper proposes an application of L-SHADE algorithm, an advanced form of Differential Evolution (DE) algorithm, to minimize the objective cost/kW. SHADE is a success history based parameter adaptation technique of DE. L-SHADE improves the performance of SHADE with linearly reducing the population size in successive generations. DE has historically been used mainly for optimization of continuous variables. The present study suggests an approach of using algorithm L-SHADE in discrete location optimization problem. Case studies of varying wind directions with constant and variable wind speeds have been performed and results are compared with some of the previous studies.

Teaching Learning Based Optimization with focused learning and its performance on CEC2017 functions

Abstract: In this work, we propose a variant to the Teaching Learning Based Optimization algorithm by incorporating focused learning of students. A student undergoes focused learning phase only when it is unable to obtain a better solution in the teacher phase and is expected to efficiently utilize the limited functional evaluations. The performance of this variant is evaluated on the single objective bound constrained real-parameter numerical optimization problems which have been proposed as a part of IEEE Congress on Evolutionary Computation. The proposed variant has provided competitive results in most of the problems.

Enhanced firefly algorithm for constrained numerical optimization

Abstract: Firefly algorithm is one of the recent and very promising swarm intelligence metaheuristics for tackling hard optimization problems. While firefly algorithm has been proven on various numerical and engineering optimization problems as a robust metaheuristic, it was not properly tested on a wide set of constrained benchmark functions. We performed testing of the original firefly algorithm on a set of standard 13 benchmark functions for constrained problems and it exhibited certain deficiencies, primarily insufficient exploration during early stage of the search. In this paper we propose enhanced firefly algorithm where main improvements are correlated to the hybridization with the exploration mechanism from another swarm intelligence algorithm, introduction of new exploitation mechanism and parameter-based tuning of the exploration-exploitation balance. We tested our approach on the same standard benchmark functions and showed that it not only overcame weaknesses of the original firefly algorithm, but also outperformed other state-of-the-art swarm intelligence algorithms.

Evolving heuristics for Dynamic Vehicle Routing with Time Windows using genetic programming

Abstract: Dynamic vehicle routing problem with time windows is an important combinatorial optimisation problem in many real-world applications. The most challenging part of the problem is to make real-time decisions (i.e. whether to accept the newly arrived service requests or not) during the execution of the routes. It is hardly applicable to use the optimisation methods such as mathematical programming and evolutionary algorithms that are competitive for static problems, since they are usually time-consuming, and cannot give real-time responses. In this paper, we consider solving this problem using heuristics. A heuristic gradually builds a solution by adding the requests to the end of the route one by one. This way, it can take advantage of the latest information when making the next decision, and give immediate response. In this paper, we propose a meta-algorithm to generate a solution given any heuristic. The meta-algorithm maintains a set of routes throughout the scheduling horizon. Whenever a new request arrives, it tries to re-generate new routes to include the new request by the heuristic. It accepts the new request if successful, and reject otherwise. Then we manually designed several heuristics, and proposed a genetic programming-based hyper-heuristic to automatically evolve heuristics. The results showed that the heuristics evolved by genetic programming significantly outperformed the manually designed heuristics.

Evaluating and modelling Hanabi-playing agents

Abstract: Agent modelling involves considering how other agents will behave, in order to influence your own actions. In this paper, we explore the use of agent modelling in the hidden-information, collaborative card game Hanabi. We implement a number of rule-based agents, both from the literature and of our own devising, in addition to an Information Set-Monte Carlo Tree Search (IS-MCTS) agent. We observe poor results from IS-MCTS, so construct a new, predictor version that uses a model of the agents with which it is paired. We observe a significant improvement in game-playing strength from this agent in comparison to IS-MCTS, resulting from its consideration of what the other agents in a game would do. In addition, we create a flawed rule-based agent to highlight the predictor's capabilities with such an agent.

Multi-method based orthogonal experimental design algorithm for solving CEC2017 competition problems

Abstract: Over the last two decades, many different evolutionary algorithms (EAs) have been proposed for solving optimization problems. However, no single EA has consistently been the best for solving a wide range of them. In the literature, this drawback has been tackled by using multiple EAs in a single framework. In this paper, a new multi-method based EA that utilizes the search ability of multi-operator differential evolution algorithm (MODE) and covariance matrix adaptation evolution strategy CMA-ES algorithm in a single framework, has been presented, with the orthogonal experimental design (OED) and factor analysis (FA) used to select the proper combination of mutation strategies, control parameters adaptation strategies, and crossover operators. To judge the performance of this algorithm, 30 problems are solved from the CEC2017 competition and their results are analyzed.

A comprehensive survey of brain storm optimization algorithms

Abstract: The development, implementation, variant, and future directions of a new swarm intelligence algorithm, brain storm optimization (BSO) algorithm, are comprehensively surveyed. Brain storm optimization algorithm is a new and promising swarm intelligence algorithm, which simulates the human brainstorming process. Through the convergent operation and divergent operation, individuals in BSO are grouped and diverged in the search space/objective space. To the best of our knowledge, there are 75 papers, 8 theses, and 5 patents in total on the development and application of the BSO algorithm. Every individual in the BSO algorithm is not only a solution to the problem to be optimized, but also a data point to reveal the landscape of the problem. Based on the developments of brain storm optimization algorithms, different kinds of optimization problems and real-world applications could be solved.

Surrogate-assisted evolutionary multiobjective shape optimization of an air intake ventilation system

Abstract: We tackle three different challenges in solving a real-world industrial problem: formulating the optimization problem, connecting different simulation tools and dealing with computationally expensive objective functions. The problem to be optimized is an air intake ventilation system of a tractor and consists of three computationally expensive objective functions. We describe the modeling of the system and its numerical evaluation with a commercial software. To obtain solutions in few function evaluations, a recently proposed surrogate-assisted evolutionary algorithm K-RVEA is applied. The diameters of four different outlets of the ventilation system are considered as decision variables. From the set of nondominated solutions generated by K-RVEA, a decision maker having substance knowledge selected the final one based on his preferences. The final selected solution has better objective function values compared to the baseline solution of the initial design. A comparison of solutions with K-RVEA and RVEA (which does not use surrogates) is also performed to show the potential of using surrogates.

The N-Tuple bandit evolutionary algorithm for automatic game improvement

Abstract: This paper describes a new evolutionary algorithm that is especially well suited to AI-Assisted Game Design. The approach adopted in this paper is to use observations of AI agents playing the game to estimate the game's quality. Some of best agents for this purpose are General Video Game AI agents, since they can be deployed directly on a new game without game-specific tuning; these agents tend to be based on stochastic algorithms which give robust but noisy results and tend to be expensive to run. This motivates the main contribution of the paper: the development of the novel N-Tuple Bandit Evolutionary Algorithm, where a model is used to estimate the fitness of unsampled points and a bandit approach is used to balance exploration and exploitation of the search space. Initial results on optimising a Space Battle game variant suggest that the algorithm offers far more robust results than the Random Mutation Hill Climber and a Biased Mutation variant, which are themselves known to offer competitive performance across a range of problems. Subjective observations are also given by human players on the nature of the evolved games, which indicate a preference towards games generated by the N-Tuple algorithm.

A note on constrained multi-objective optimization benchmark problems

Abstract: We investigate the properties of widely used constrained multi-objective optimization benchmark problems. A number of Multi-Objective Evolutionary Algorithms (MOEAs) for Constrained Multi-Objective Optimization Problems (CMOPs) have been proposed in the past few years. The C-DTLZ functions and Real-World-Like Problems (RWLPs) have frequently been used for evaluating the performance of MOEAs on CMOPs. In this paper, however, we show that the C-DTLZ functions and widely-used RWLPs have some unnatural problem features. The experimental results show that an MOEA without any Constraint Handling Techniques (CHTs) can successfully find well-approximated nondominated feasible solutions on the C1-DTLZ1, C1-DTLZ3, and C2-DTLZ2 functions. It is widely believed that RWLPs are MOEA-hard problems, and finding the feasible solutions on them is a very hard task. However, we show that the MOEA without any CHTs can find feasible solutions on widely-used RWLPs such as the speed reducer design problem, the two-bar truss design problem, and the water problem. Also, it is seldom that the infeasible solution simultaneously violates multiple constraints in the RWLPs. Due to the above reasons, we conclude that constrained multi-objective optimization benchmark problems need a careful reconsideration.

Population seeding techniques for Rolling Horizon Evolution in General Video Game Playing

Abstract: While Monte Carlo Tree Search and closely related methods have dominated General Video Game Playing, recent research has demonstrated the promise of Rolling Horizon Evolutionary Algorithms as an interesting alternative. However, there is little attention paid to population initialization techniques in the setting of general real-time video games. Therefore, this paper proposes the use of population seeding to improve the performance of Rolling Horizon Evolution and presents the results of two methods, One Step Look Ahead and Monte Carlo Tree Search, tested on 20 games of the General Video Game AI corpus with multiple evolution parameter values (population size and individual length). An in-depth analysis is carried out between the results of the seeding methods and the vanilla Rolling Horizon Evolution. In addition, the paper presents a comparison to a Monte Carlo Tree Search algorithm. The results are promising, with seeding able to boost performance significantly over baseline evolution and even match the high level of play obtained by the Monte Carlo Tree Search.

Common subtrees in related problems: A novel transfer learning approach for genetic programming

Abstract: Transfer learning is a machine learning technique which has demonstrated great success in improving outcomes on a broad range of problems. However prior methods of transfer learning in Genetic Programming (GP) have tended to rely on random processes or meta-knowledge of the problem structure to facilitate selection of information for use in transfer. To address these issues, a non-random method for automatically finding relevant information for transfer between two source domain problems from the same problem domain based on common subtrees is proposed. This information is then utilised within a modular transfer learning framework, being added to the function set for a target problem prior to population initialisation. The performance of the proposed method is assessed using multiple benchmark problems from two distinct problem domains, namely symbolic regression and Boolean domain problems, and compared to standard GP and the-state-of-the-art transfer learning method for the given problems. The results show that the newly introduced method has either significantly outperformed, or achieved comparable performance to, the competitor methods on the problems of the two domains. We conclude that the proposed method demonstrates ability as a general transfer learning technique for GP and note some possible avenues for future research based off these results.

A novel approach for Hepatocellular Carcinoma detection and classification based on triphasic CT Protocol

Abstract: Introduction and objective: Computer Aided Decision (CAD) systems based on Medical Imaging could support radiologists in grading Hepatocellular carcinoma (HCC) by means of Computed Tomography (CT) images, avoiding medical invasive procedures such as biopsies. The identification and characterization of Regions of Interest (ROIs) containing lesions is an important phase allowing an easier classification in two classes of HCCs. Two steps are needed for the detection of lesioned ROIs: a liver isolation in each CT slice and a lesion segmentation.

Evolving Deep Neural Networks architectures for Android malware classification

Abstract: Deep Neural Networks (DNN) have become a powerful, widely used, and successful mechanism to solve problems of different nature and varied complexity. Their ability to build models adapted to complex non-linear problems, have made them a technique widely applied and studied. One of the fields where this technique is currently being applied is in the malware classification problem. The malware classification problem has an increasing complexity, due to the growing number of features needed to represent the behaviour of the application as exhaustively as possible. Although other classification methods, as those based on SVM, have been traditionally used, the DNN pose a promising tool in this field. However, the parameters and architecture setting of these DNNs present a serious restriction, due to the necessary time to find the most appropriate configuration. This paper proposes a new genetic algorithm designed to evolve the parameters, and the architecture, of a DNN with the goal of maximising the malware classification accuracy, and minimizing the complexity of the model. This model is tested against a dataset of malware samples, which are represented using a set of static features, so the DNN has been trained to perform a static malware classification task. The experiments carried out using this dataset show that the genetic algorithm is able to select the parameters and the DNN architecture settings, achieving a 91% accuracy.