Algorithms using swarming collective motion can solve coverage problems in unknown environments by reacting to unknown obstacles in real-time when they are encountered. However, these algorithms face two key challenges when deployed on real robots. First, hand-tuning efficient collective motion parameters is both time-consuming and difficult. Second, predicting the time required for a swarm to solve a particular problem is not straightforward. This paper introduces a novel evolutionary framework to address both problems by proposing a methodology that autonomously tunes collective motion parameters for coverage problems while predicting the time required for real robots to complete the task. Our approach utilizes a simulation–optimization framework that employs a genetic algorithm to optimize the parameters of a frontier-led swarming algorithm. Results indicate that the optimized parameters are transferable to real robots, achieving 100% coverage while maintaining 84% connectivity between them. Compared to state-of-the-art swarm methods, our system reduced turnaround time by 50% and 57% in different environments while maintaining collective motion. It also achieved a 55% reduction in turnaround time on average across five scenarios compared to budget-constrained path planning, with a 10% increase in coverage. Furthermore, our framework outperformed both hand-tuned and learned collective motion approaches, reducing turnaround time by 73% in non-collective motion scenarios and by 63% while maintaining 85% connectivity in collective motion scenarios. This approach effectively combines the adaptability of swarm behavior with the predictive reliability of planning methods. • Proposed a framework for optimizing frontier-led swarming parameters. • Utilized a genetic algorithm to enhance swarm robot performance in coverage tasks. • Developed a dynamic swarm simulator for a seamless transition to physical robots. • Evaluated swarm performance metrics balancing connectivity and coverage time. • Results highlight the effectiveness of optimized parameters in real-world scenarios.

Recognition of swarm behavior is important for two reasons. First, it permits the early detection of adversarial collective motion behaviours such that counter-collective motion can be activated. Second, it permits the monitoring and assessment of own swarms to enable detecting any disruptions to the required behavior. Existing work in this area requires feature-based data provided by an external observer that has access to all the swarm states that could not be available all the time. However, the need for pre-processing swarm data to calculate its features can lead to inefficient behavior recognition. This paper addresses this limitation by using raw video data for swarm behavior recognition. This paper proposes a new framework to autonomously recognize structured collective behavior in swarm robots from camera data First, we present a dataset of both collective motion and random behaviors of Pioneer 3DX robots. Then we formulate the recognition problem as a spatiotemporal learning problem. A recurrent neural network is used as the main building block to evaluate each video representation—Our experimental results showed that this video-based recognition without any data pre-processing results in high accuracy that is comparative to feature-based recognition techniques. The proposed model is able to efficiently recognise robots’ behaviour with 99.8% accuracy. Also, our methodology can distinguish behaviours of real robots with 79% accuracy even with different numbers of agents.

This paper proposes an innovative approach to coverage path planning and obstacle avoidance for multiple Unmanned Ground Vehicles (UGVs) in a changing environment, taking into account constraints on the time, path length, number of UGVs and obstacles. Our approach leverages deformable virtual leader-follower formations to enable UGVs to adapt their formation based on both planned and real-time sensor data. A hierarchical block algorithm is employed to identify areas in the environment where UGV formations can spread out to meet time and budget constraints. Additionally, we introduce a novel control scheme that allows each UGV to generate a local steering force to dodge any static and mobile obstacles based on the closest safe angle. Results from simulations and real UGV experiments demonstrate that our approach achieves a higher coverage percentage than rule-based and reactive swarming approaches without planning. Our approach offers a promising solution for efficient coverage path planning and obstacle avoidance in complex environments with multiple UGVs.

Swarming robots have the potential to perform many different tasks like coverage, exploration, and navigation in industry, healthcare, military and transportation. However, human control of large numbers of robots is difficult. It is not yet clear which metrics may be beneficial for human-swarm interaction or how to display them. This paper presents an Android platform for permitting human-swarm interaction, while also displaying metrics describing the swarm behavior. The visualization includes charts to represent the changes in boids' grouping, alignment, fragmentation, and coverage metrics. Experiments show how the visualization responds to different behaviors of swarm triggered by human interaction.

A single dataset could hide a significant number of relationships among its feature set. Learning these relationships simultaneously avoids the time complexity associated with running the learning algorithm for every possible relationship, and affords the learner with an ability to recover missing data and substitute erroneous ones by using available data. In our previous research, we introduced the gate-layer autoencoders (GLAEs), which offer an architecture that enables a single model to approximate multiple relationships simultaneously. GLAE controls what an autoencoder learns in a time series by switching on and off certain input gates, thus, allowing and disallowing the data to flow through the network to increase network's robustness. However, GLAE is limited to binary gates. In this article, we generalize the architecture to weighted gate layer autoencoders (WGLAE) through the addition of a weight layer to update the error according to which variables are more critical and to encourage the network to learn these variables. This new weight layer can also be used as an output gate and uses additional control parameters to afford the network with abilities to represent different models that can learn through gating the inputs. We compare the architecture against similar architectures in the literature and demonstrate that the proposed architecture produces more robust autoencoders with the ability to reconstruct both incomplete synthetic and real data with high accuracy.

Individual differences in motivation can explain why people act differently in the same situation, and which aspects of a game people with different motive profiles may find most engaging. However, identifying a player’s motive profile from data available during gameplay remains an open research question. Besides a range of subjective and objective techniques for identifying player motivation, electroencephalography (EEG) technology could offer an automatic, objective technique for identifying the profile that best describes a given player. This article proposes a framework to measure player profiles of achievement, affiliation, and power motivation using EEG signals during their engagement within a game. First, an abstract mini-game is proposed to evaluate a player’s motivation. In the mini-game, each human player interacts with four nonplayer characters to gainfortuneorfriendshipthrough an individual play phase and a social network phase. The game is used within an experimental scenario to collect players’ actions and EEG signals. In addition, data from a psychological test are used to establish ground truth. We propose three subject labeling schemes using the output of the psychological test. Based on a player’s motive profile, behavioral indicators and EEG data analysis indicate that assessing a player’s motive profile is more robust from EEG signals than from behavioral data.

This paper presents a grammar-based evolutionary model to facilitate autonomous emergence of task allocation for intelligent multi-agent systems. The approach adopts a context-free grammar to determine the behaviour rule syntax. This allows for flexibility in evolving task allocation under multiple and dynamic constraints without manual rule design and parameter tuning. Experimental evaluations conducted with a target discovery simulation illustrate that the grammar-based model performs successfully in both dynamic and non-dynamic conditions. A statistically significant performance improvement is shown compared to an algorithm developed with the broadcast of local eligibility mechanism and a genetic programming mechanism. Grammatical evolution can achieve near-optimal solutions under restrictions applied on the number of agents, targets and the time allowed. Further, analysis of the evolved rule structures shows that grammatical evolution can identify less complex rule structures for behaviours while maintaining the expected level of performance. The results infer that the proposed model is a promising alternative for dynamic task allocation with human interactions in complex real-world domains.

This paper presents a grammar-based evolutionary approach that incorporates abstractions to learn complex collective behaviours through their simpler representations. We propose modifications to the grammar syntax design and genome structure to facilitate evolution of abstractions in separate genome partitions. Two abstraction techniques based on behavioural decomposition and environmental scaffolding are presented to derive these simpler representations. Parallel and incremental learning architectures incorporated with grammatical evolution (GE) are investigated with three complex problems to evaluate their potential in generating collective multi-agent behaviours. The results infer that both learning architectures surpass a generic GE model in performance for evolving complex behaviours. Furthermore, using environmental scaffolding reduces the robustness of the model than when only the behavioural decomposition technique is used. However, it has more potential to generate solutions with better fitness than when scaffoldings are not used. The evaluations suggest that, by incorporating abstraction learning architectures with grammar-based evolution can significantly improve the performance of an agent system in complex problem domains.

Trust has been widely recognized as particularly significant among the factors influencing team performance. Trust directly impacts team performance as it plays a pivotal role in the decisions that each team member (each agent) makes regarding their own actions and their interactions with fellow team members (other agents). Hence, determining the appropriate actions of each agent in the team based on perceived trust information is critical to ensure optimal team performance. However, there is no generic mechanism dedicated to such a problem. This paper addresses this problem by proposing a novel trust architecture which integrates differentiated trust with response strategies. Differentiated trust is multidimensional trust with each trust dimension representing trust in an agent's abilities to perform the task associated with that dimension. With differentiated trust, an agent can differentiate the trustworthiness of another agent in performing different subtasks (a secondary task or a portion of the primary task). To further fulfill the transition from perceptual trust to practical actions, responses strategies are introduced. Each response strategy associates trust levels with the available actions through a distinct deterministic strategy. The high dimensional trust enabled by the differentiated trust is used as the input of a response strategy for more nuanced manipulation of the interactions. The impact of the proposed trust architecture is demonstrated through an experimental investigation. A platform simulating team performance is built based on a food foraging task. Scenarios embodying different types and proportions of a priori flawed agents are introduced to enable the system to distinguish between agents in terms of trust; thus, providing a potential to optimize team performance through the trust architecture. It is demonstrated that the proposed trust architecture enables optimized team performance in various scenarios involving agents with different trustworthiness by the appropriate determination of each agent's actions in the interactive teamwork.

Considerable progress has been made in improving the estimation accuracy of cognitive workload using various sensor technologies. However, the overall performance of different algorithms and methods remain suboptimal in real-world applications. Some studies in the literature demonstrate that a single modality is sufficient to estimate cognitive workload. These studies are limited to controlled settings, a scenario that is significantly different from the real world where data gets corrupted, interrupted, and delayed. In such situations, the use of multiple modalities is needed. Multimodal fusion approaches have been successful in other domains, such as wireless-sensor networks, in addressing single-sensor weaknesses and improving information quality/accuracy. These approaches are inherently more reliable when a data source is lost. In the cognitive workload literature, sensors, such as electroencephalography (EEG), electrocardiography (ECG), and eye tracking, have shown success in estimating the aspects of cognitive workload. Multimodal approaches that combine data from several sensors together can be more robust for real-time measurement of cognitive workload. In this article, we review the published studies related to multimodal data fusion to estimate the cognitive workload and synthesize their main findings. We identify the opportunities for designing better multimodal fusion systems for cognitive workload modeling.

Differential evolution was mainly proposed for solving optimization problems with continuous decision variables because of its Euclidean distance-based learning concept. This made it unsuitable for many binary and discrete problems. However, several studies approved the applicability of differential evolution algorithm for effectively solving such problems. In this paper, a new design of differential evolution, which incorporates mapping and repairing methods, modified mutation operator and local searches, is proposed to solve the complex multicomponents traveling thief problems that are characterized by both binary and discrete parameters. Also, a novel initialization and repairing method, which enables differential evolution's operators to only evolve solutions of one component and optimally distribute/update the solutions of the other one with considering the inter-dependency between both components, is introduced. To judge the performance of the proposed algorithm, 13 strongly correlated instances of traveling thief problems have been solved and the results have been compared with those from 24 selfdesigned and state-of-the-art algorithms. Results demonstrated the competitive performance of the proposed algorithm in terms of the quality of obtained solutions and computational time.

Background: Using design metrics to predict fault-prone elements of a software design can help to focus attention on classes that need redesign and more extensive testing. However, some design metrics have been pointed out to be theoretically invalid, and the usefulness of some metrics is questioned.

Swarm guidance, such as the case of guiding a group of sheep away from a field, is a challenging task. As the swarm size increases, it becomes necessary that multiple control points, or sheepdogs, are needed to guide the swarm. In this paper, a swarm of unmanned aerial vehicles (UAVs) acts as a moving safety network (aka a formation) that not only guides the sheep swarm, but also prevents them from dispersing or reversing to the other side of the field. We investigate two types of formations. The first type acts as a baseline, maintains fixed distances from the sheep swarm, and relies on fixed predefined angular structure relative to the sheep's global centre of mass (GCM). The second type is dynamic, where the force vector to control the UAV and the individual distance of each UAV from the sheep's GCM are controlled by a Perceptron, with the weights optimized by a particle swarm optimization algorithm. We evolve five Perceptrons to specialize in relative positions in the formation, which fixes the space cost for the optimization algorithm, while allowing the size of the swarm of UAVs to scale up. We demonstrate that the use of Perceptron-networks for dynamic control scheme reduces the total distance travelled by the UAVs, is transparent when interpreted with Hinton diagrams, and transferable to a larger number of UAVs.

With the continuous technological advancement in the game industry, millions of players engage in various online games everyday. Player population size of games ebb and flow through time as a complex series. Analyzing these player population numbers in a shorter time window, such as weekly, could help generate insights that enrich the understanding about low-level population fluctuation patterns of online games. However, this area of game data analytics still has space for further enhancement. This study focuses on discovering patterns of weekly player population fluctuations, that could aid in comprehending how the population of various kind of games change within a framing window of a week. We use player population time series of 1963 games available on Steam. Utilizing several trend removal techniques and conducting seasonality detection we identify that 77% of games display a recurring weekly pattern in player population fluctuations. Moreover, our dynamic time warping based cluster analysis shows that there are 9 diverse weekly player population fluctuation patterns. Among these 9, the governing pattern visible in the majority of games displays that the player population is higher towards the weekend. Finally, we scrutinize the tags, age requirements and overall population size of games in each cluster associated with the diverse patterns to generate insights about the characteristics of games associated with each weekly population pattern.

The traveling salesman problem is a well-known combinatorial optimization problem with permutation-based variables, which has been proven to be an NP-complete problem. Over the last few decades, many evolutionary algorithms have been developed for solving it. In this study, a new design that uses the k-means clustering method, is proposed to be used as a repairing method for the individuals in the initial population. In addition, a new crossover operator is introduced to improve the evolving process of an evolutionary algorithm and hence its performance. To investigate the performance of the proposed mechanism, two popular evolutionary algorithms (genetic algorithm and differential evolution) have been implemented for solving 18 instances of traveling salesman problems and the results have been compared with those obtained from standard versions of GA and DE, and 3 other state-of-the-art algorithms. Results show that the proposed components can significantly improve the performance of EAs while solving TSPs with small, medium and large-sized problems.

Autoencoders (AE) have been used successfully as unsupervised learners for inferring latent information, learning hidden features and reducing the dimensionality of the data. In this paper, we propose a new AE architecture: Gate-Layer AE (GLAE). The novelty of GLAE lies in its ability to encourage learning of the relationships among different input variables, which affords it with an inherent ability to recover missing variables from the available ones and to act as a concurrent multi-function approximator.GLAE uses a network architecture that associates each input with a binary gate acting as a switch that turns on or off the flow to each input unit, while synchronising its action with data flow to the network. We test GLAE with different coding sizes and compare its performance against the Classic AE, Denoising AE and Variational AE. The evaluation uses Electroencephalograph (EEG) data with an aim to reconstruct the EEG signal when some data are missing. The results demonstrate GLAE's superior performance in reconstructing EEG signals with up to 25% missing data in an input stream.

The traditional differential evolution algorithm was originally, and still is mainly, used to solve continuous optimization problems. As a result, it has not commonly been considered as applicable for several real-world problems in the permutation-based domain. In this paper, a novel differential evolution algorithm, which incorporates several effective components, is introduced. These components increase search effectiveness by providing a good balance between exploration (discovering new solutions) and exploitation (further exploring current solutions) processes. Moreover, a dual representation of solutions, which has the capability to allow normal continuous handling of variables by differential evolution operators, and at same time provide binary variables for fitness measurement, is employed. To judge the performance of the proposed algorithm, 14 instances of 0-1 knapsack problems have been solved and the results have been compared with those obtained from 11 state-of-the-art algorithms. Results show that the proposed algorithm was able to outperform other algorithms in solving small and medium sized knapsack problems and is competitive in large-sized problems.

This paper presents an evolutionary computing based approach to automatically synthesise swarm behavioural rules from their atomic components, thus making a step forward in trying to mitigate human bias from the rule generation process, and leverage the full potential of swarm systems in the real world by modelling more complex behaviours. We identify four components that make-up the structure of a rule: control structures, parameters, logical/relational connectives and preliminary actions, which form the rule space for the proposed approach. A boids simulation system is employed to evaluate the approach with grammatical evolution and genetic programming techniques using the rule space determined. While statistical analysis of the results demonstrates that both methods successfully evolve desired complex behaviours from their atomic components, the grammatical evolution model shows more potential in generating complex behaviours in a modularised approach. Furthermore, an analysis of the structure of the evolved rules implies that the genetic programming approach only derives non-reusable rules composed of a group of actions that is combined to result in emergent behaviour. In contrast, the grammatical evolution approach synthesises sound and stable behavioural rules which can be extracted and reused, hence making it applicable in complex application domains where manual design is infeasible.

A formation is a geometric shape that a group of agents spatially organizes themselves into and maintains over time. Swarm Q-Learning (SQL) is a tabular multi-agent reinforcement learning algorithm designed to solve formation control problems. We modify SQL by allowing agents to exchange knowledge they have learnt within the same environment and introduce the Swarm Q-Learning with knowledge Sharing within an Environment (SQL-SIE). The algorithm is tested on a task where a swarm of robots, initially scattered in one side of the environment, needs to navigate through obstacles until they reach their initial positions in the formation within a region of interest. Experimental results show that the proposed SQL-SIE is more efficient than SQL as measured by the time taken by the swarm to complete this part of the mission. Moreover, SQL-SIE scales better than SQL as the number of agents increases.

EEG signals could reveal unique information of an individual's brain activities. They have been regarded as one of the most promising biometric signals for person identification and verification. Steady-State Visual Evoked Potentials (SSVEPs), as EEG responses to visual stimulations at specific frequencies, could provide biometric information. However, current methods on SSVEP biometrics with hand-crafted power spectrum features and canonical correlation analysis (CCA) present only a limited range of individual distinctions and suffer relatively low accuracy. In this paper, we propose convolution neural networks (CNNs) with raw SSVEPs for person identification and verification without the need for any hand-crafted features. We conduct a comprehensive comparison between the performance of CNN with raw signals and a number of classical methods on two SSVEP datasets consisting of four and ten subjects, respectively. The proposed method achieved an averaged identification accuracy of 96.8%±0.01, which outperformed the other methods by an average of 45.5% (p-value <; 0.05). In addition, it achieved an averaged False Acceptance Rate (FAR) of 1.53%±0.01 and True Acceptance Rate (TAR) of 97.09%±0.02 for person verification. The averaged verification accuracy is 98.34% ± 0.01, which outperformed the other methods by an average of 11.8% (p-value <; 0.05). The proposed method based on deep learning offers opportunities to design a general-purpose EEG-based biometric system without the need for complex pre-processing and feature extraction techniques, making it feasible for real-time embedded systems.

Robotic control in a continuous action space has long been a challenging topic. This is especially true when controlling robots to solve compound tasks, as both basic skills and compound skills need to be learned. In this paper, we propose a hierarchical deep reinforcement learning algorithm to learn basic skills and compound skills simultaneously. In the proposed algorithm, compound skills and basic skills are learned by two levels of hierarchy. In the first level of hierarchy, each basic skill is handled by its own actor, overseen by a shared basic critic. Then, in the second level of hierarchy, compound skills are learned by a meta critic by reusing basic skills. The proposed algorithm was evaluated on a Pioneer 3AT robot in three different navigation scenarios with fully observable tasks. The simulations were built in Gazebo 2 in a robot operating system Indigo environment. The results show that the proposed algorithm can learn both high performance basic skills and compound skills through the same learning process. The compound skills learned outperform those learned by a discrete action space deep reinforcement learning algorithm.

In this paper, we propose a deep reinforcement learning algorithm to learn multiple tasks concurrently. A new network architecture is proposed in the algorithm which reduces the number of parameters needed by more than 75% per task compared to typical single-task deep reinforcement learning algorithms. The proposed algorithm and network fuse images with sensor data and were tested with up to 12 movement-based control tasks on a simulated Pioneer 3AT robot equipped with a camera and range sensors. Results show that the proposed algorithm and network can learn skills that are as good as the skills learned by a comparable single-task learning algorithm. Results also show that learning performance is consistent even when the number of tasks and the number of constraints on the tasks increased.

Ensemble methods aim at combining multiple learning machines to improve the efficacy in a learning task in terms of prediction accuracy, scalability, and other measures. These methods have been applied to evolutionary machine learning techniques including learning classifier systems (LCSs). In this article, we first propose a conceptual framework that allows us to appropriately categorize ensemble‐based methods for fair comparison and highlights the gaps in the corresponding literature. The framework is generic and consists of three sequential stages: a pre‐gate stage concerned with data preparation; the member stage to account for the types of learning machines used to build the ensemble; and a post‐gate stage concerned with the methods to combine ensemble output. A taxonomy of LCSs‐based ensembles is then presented using this framework. The article then focuses on comparing LCS ensembles that use feature selection in the pre‐gate stage. An evaluation methodology is proposed to systematically analyze the performance of these methods. Specifically, random feature sampling and rough set feature selection‐based LCS ensemble methods are compared. Experimental results show that the rough set‐based approach performs significantly better than the random subspace method in terms of classification accuracy in problems with high numbers of irrelevant features. The performance of the two approaches are comparable in problems with high numbers of redundant features.

In this paper, we introduce a novel form of association rules (ARs) that do not require discretization of continuous variables or the use of intervals in either sides of the rule. This rule form captures nonlinear relationships among variables, and provides an alternative pattern representation for mining essential relations hidden in a given data set. We refer to the new rule form as a functional AR (FAR). A new neural network-based, co-operative, coevolutionary algorithm is presented for FAR mining. The algorithm is applied to both synthetic and real-world data sets, and its performance is analyzed. The experimental results show that the proposed mining algorithm is able to discover valid and essential underlying relations in the data. Comparison experiments are also carried out with the two state-of-the-art AR mining algorithms that can handle continuous variables to demonstrate the competitive performance of the proposed method.

Classification rule induction involves two main processes: finding the optimal conjuncts (attribute intervals or attribute-value pairs) and their combination (disjuncts or rules) to classify different concepts in the data. The evolutionary rule learning approaches employ an evolutionary algorithm, such as a genetic algorithm, to perform both these search operations simultaneously. This approach often leads to significant problems including population bloating and stalled evolutionary search in real-valued attribute problems, especially with higher dimensions. In this paper, we present an online evolutionary rule learning approach referred to as ERL-AID that decouples the above search processes and employs a discretization algorithm that works on the attribute space and a genetic algorithm to combine the discretized attributes into appropriate classification rules. ERL-AID applies a sliding window approach to process inputs in an online fashion. The proposed system is able to produce compact rule sets with competitive performance and could scale to higher dimensions. The experimental results show the competitiveness of our algorithm.

This paper considers task allocation problems where a group of agents must discover and allocate themselves to tasks. Task allocation is particularly difficult when agents can only exchange information over a limited communication range and when the agents are initialized from a single departure point. To address these constraints, we present a novel approach that incorporates computational models of motivation into a guaranteed convergence particle swarm optimization algorithm. We introduce an incentive function and three motive profiles to guaranteed convergence particle swarm optimization. Our new algorithm is compared to existing approaches with and without motivation under conditions of limited communication. It is tested in the case where the agents are initialized from a single point and random points. Results show that our approach increases the number of tasks discovered by a group of agents under these conditions. Furthermore, it significantly outperforms benchmark PSO algorithms in the number of tasks discovered and allocated when the agents are initialized from a single point.

The paper considers the task allocation problem in the case where there is a small number of agents initialized at a single point. The objective is to achieve an even distribution of agents to tasks. To address this problem, this paper proposes a new method that endows agents with models of motivation and leadership to aid their coordination. The proposed approach uses the Particle Swarm Optimization algorithm with a ring neighborhood topology as a foundation and incorporates computational models of motivation to achieve the goals of task allocation more effectively. Simulation results show that, first, the proposed method increases the number of tasks discovered. Secondly, the number of tasks to which the agents are allocated increases. Thirdly, the agents distribute themselves more evenly among the tasks.

Agents are systems capable of perceiving their environment through sensors, reasoning about their sensory input using some characteristic reasoning process and acting in their environment using effectors. When one or more agents control the objects that comprise a 3D virtual world, the result is a dynamic, adaptive environment that changes in response to users' actions. We have experimented with three different agent models for this purpose: a swarm model, a cognitive model and a motivated agent model. Each of these models differs in the complexity of its implementation and can thus be used to produce dynamic virtual environments of differing behavioural complexity. This paper introduces a schema for characterising the implementation and behavioural complexity of agent models for dynamic virtual environments. We apply this schema to the agent models we have studied to reveal their advantages and disadvantages and identify directions for future work.