Emma Hart

Edinburgh Napier University · School of Computing

The discrepancy between simulated and hardware experiments, the reality gap, is a challenge in evolutionary robotics. While strategies have been proposed to address this gap in fixed-body robots, they are not viable when dealing with populations and generations where the body is in constant change. The continual evolution of body designs necessitat...

Deep neural networks (DNN) are increasingly being used to perform algorithm-selection in combinatorial optimisation domains, particularly as they accommodate input representations which avoid designing and calculating features. Mounting evidence from domains that use images as input shows that deep convolutional networks are vulnerable to adversari...

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 resp...

Lengthy evaluation times are common in many optimization problems such as direct policy search tasks, especially when they involve conducting evaluations in the physical world, e.g. in robotics applications. Often when evaluating solution over a fixed time period it becomes clear that the objective value will not increase with additional computatio...

The evolutionary robotics field offers the possibility of autonomously generating robots that are adapted to desired tasks by iteratively optimising across successive generations of robots with varying configurations until a high-performing candidate is found. The prohibitive time and cost of actually building this many robots means that most evolu...

Lengthy evaluation times are common in many optimization problems such as direct policy search tasks, especially when they involve conducting evaluations in the physical world, e.g. in robotics applications. Often, when evaluating a solution over a fixed time period, it becomes clear that the objective value will not increase with additional comput...

To advance research in the development of optimisation algorithms, it is crucial to have access to large test-beds of diverse and discriminatory instances from a domain that can highlight strengths and weaknesses of different algorithms. The DIGNEA tool enables diverse instance suites to be generated for any domain, that are also discriminatory wit...

Dynamic algorithm selection aims to exploit the complementarity of multiple optimization algorithms by switching between them during the search. While these kinds of dynamic algorithms have been shown to have potential to outperform their component algorithms, it is still unclear how this potential can best be realized. One promising approach is to...

We propose a novel technique for algorithm-selection, applicable to optimisation domains in which there is implicit sequential information encapsulated in the data, e.g., in online bin-packing. Specifically we train two types of recurrent neural networks to predict a packing heuristic in online bin-packing, selecting from four well-known heuristics...

Jointly optimising both the body and brain of a robot is known to be a challenging task, especially when attempting to evolve designs in simulation that will subsequently be built in the real world. To address this, it is increasingly common to combine evolution with a learning algorithm that can either improve the inherited controllers of new offs...

We propose two new methods for evolving the layout of an instance-space. Specifically we design three different fitness metrics that seek to: (i) reward layouts which place instances won by the same solver close in the space; (ii) reward layouts that place instances won by the same solver and where the solver has similar performance close together;...

Using Neuroevolution combined with Novelty Search to promote behavioural diversity is capable of constructing high-performing ensembles for classification. However, using gradient descent to train evolved architectures during the search can be computationally prohibitive. Here we propose a method to overcome this limitation by using a surrogate mod...

We propose a novel technique for algorithm-selection, applicable to optimisation domains in which there is implicit sequential information encapsulated in the data, e.g., in online bin-packing. Specifically we train two types of recurrent neural networks to predict a packing heuristic in online bin-packing, selecting from four well-known heuristics...

Using Neuroevolution combined with Novelty Search to promote behavioural diversity is capable of constructing high-performing ensembles for classification. However, using gradient descent to train evolved architectures during the search can be computationally prohibitive. Here we propose a method to overcome this limitation by using a surrogate mod...

The joint optimisation of body-plan and control via evolutionary processes can be challenging in rich morphological spaces in which offspring can have body-plans that are very different from either of their parents. This causes a potential mismatch between the structure of an inherited controller and the new body. To address this, we propose a fram...

We survey and reflect on how learning (in the form of individual learning and/or culture) can augment evolutionary approaches to the joint optimization of the body and control of a robot. We focus on a class of applications where the goal is to evolve the body and brain of a single robot to optimize performance on a specified task. The review is gr...

Hyper-heuristic frameworks, although intended to be cross-domain at the highest level, usually rely on a set of domain-specific low-level heuristics which exist below the domain-barrier and are manipulated by the hyper-heuristic itself. However, for some domains, the number of available heuristics can be very low, while for novel problems, no heuri...

Both algorithm-selection methods and hyper-heuristic methods rely on a pool of complementary heuristics. Improving the pool with new heuristics can improve performance, however, designing new heuristics can be challenging. Methods such as genetic programming have proved successful in automating this process in the past. Typically, these make use of...

In evolutionary robotics, several approaches have been shown to be capable of the joint optimisation of body-plans and controllers by either using only evolution or combining evolution and learning. When working in rich morphological spaces, it is common for offspring to have body-plans that are very different from either of their parents, which ca...

Quality-diversity (QD) algorithms that return a large archive of elite solutions to a problem provide insights into how high-performing solutions are distributed throughout a feature-space defined by a user. They are often described as illuminating the feature-space, providing a qualitative illustration of relationships between features and objecti...

In order to address scalability issues, which can be a challenge for Deep Learning methods, we propose Wide Learning of Diverse Architectures—a model that scales horizontally rather than vertically, enabling distributed learning. We propose a distributed version of a quality-diversity evolutionary algorithm (MAP-Elites) to evolve an architecturally...

A long-term vision of evolutionary robotics is a technology enabling the evolution of entire autonomous robotic ecosystems that live and work for long periods in challenging and dynamic environments without the need for direct human oversight. Evolutionary Robotics has been widely used due to its capability of creating unique robot designs in simul...

Neural Referring Expression Generation (REG) models have shown promising results in generating expressions which uniquely describe visual objects. However, current REG models still lack the ability to produce diverse and unambiguous referring expressions (REs). To address the lack of diversity, we propose generating a set of diverse REs, rather tha...

We outline a perspective on the future of evolutionary robotics and discuss a long-term vision regarding robots that evolve in the real world. We argue that such systems offer significant potential for advancing both science and engineering. For science, evolving robots can be used to investigate fundamental issues about evolution and the emergence...

The long term goal of the Autonomous Robot Evolution (ARE) project is to create populations of physical robots, in which both the controllers and body plans are evolved. The transition for evolutionary designs from purely simulation environments into the real world creates the possibility for new types of system able to adapt to unknown and changin...

Advances in rapid prototyping have opened up new avenues of research within Evolutionary Robotics in which not only controllers but also the body plans (morphologies) of robots can evolve in real-time and real-space. However, this also introduces new challenges, in that robot models that can be instantiated from an encoding in simulation might not...

In the field of combinatorial optimisation, per-instance algorithm selection still remains a challenging problem, particularly with respect to streaming problems such as packing or scheduling. Typical approaches involve training a model to predict the best algorithm based on features extracted from the data, which is well known to be a difficult ta...

A manifest trend is that larger and more productive human groups shift from distributed to centralized decision-making. Voluntary theories propose that human groups shift to hierarchy to limit scalar stress, i.e. the increase in cost of organization as a group grows. Yet, this hypothesis lacks a mechanistic model to investigate the organizational a...

In the field of metamorphic malware detection, training a detection model with malware samples that reflect potential mutants of the malware is crucial in developing a model resistant to future attacks. In this paper, we use a Multi-dimensional Archive of Phenotypic Elites (MAP-Elites) algorithm to generate a large set of novel, malicious mutants t...

Algorithms such as MAP-Elites provide a means of allowing users to explore a solution space by returning an archive of high-performing solutions. Such an archive, can allow the user an overview of the solution space which may be useful when formulating policy around the problem itself. The number of solutions that can potentially be returned by MAP...

The success of search-based optimisation algorithms depends on appropriately balancing exploration and exploitation mechanisms during the course of the search. We introduce a mechanism that can be used with Differential Evolution (de) algorithms to adaptively manage the balance between the diversification and intensification phases, depending on cu...

The ability to detect metamorphic malware has generated significant research interest over recent years, particularly given its proliferation on mobile devices. Such malware is particularly hard to detect via signature-based intrusion detection systems due to its ability to change its code over time. This article describes a novel framework which g...

The ability to detect metamorphic malware has generated significant research interest over recent years, particularly given its proliferation on mobile devices. Such malware is particularly hard to detect via signature-based intrusion detection systems due to its ability to change its code over time. This article describes a novel framework which g...

Community energy systems (CESs) are shared energy systems in which multiple communities generate and consume energy from renewable resources. At regular time intervals, each participating community decides whether to self-supply, store, trade, or sell their energy to others in the scheme or back to the grid according to a predefined policy which al...

An increasing emphasis on reducing pollution and congestion in city centres combined with an increase in online shopping is changing the ways in which logistics companies address vehicle routing problems (VRP). We introduce the micro-depot-VRP, in which a single supply vehicle is used to supply a set of micro-depots distributed across a city; deliv...

We propose a novel technique for algorithm-selection which adopts a deep-learning approach, specifically a Recurrent-Neural Network with Long-Short-Term-Memory (RNN-LSTM). In contrast to the majority of work in algorithm-selection, the approach does not need any features to be extracted from the data but instead relies on the temporal data sequence...

To investigate how encodings influence evolving the morphology and control of modular robots, we compared three encodings: a direct encoding and two generative encodings---a compositional pattern producing network (CPPN) and a Lindenmayer System (L-System). The evolutionary progression and final performance of the direct encoding and the L-System w...

Human social hierarchy has the unique characteristic of existing in two forms. Firstly, as an informal hierarchy where leaders and followers are implicitly defined by their personal characteristics, and secondly, as an institutional hierarchy where leaders and followers are explicitly appointed by group decision. Although both forms can reduce the...

Quality-diversity algorithms such as MAP-Elites provide a means of supporting the users when finding and choosing solutions to a problem by returning a set of solutions which are diverse according to set of user-defined features. The number of solutions that can potentially be returned by MAP-Elites is controlled by a parameter that discretises the...

We describe a new rich VRP model that captures many real-world constraints, following a recently proposed taxonomy that addresses both scenario and problem physical characteristics. The model is used to generate 4800 new instances of rich VRPs which is made freely available. To the best of our knowledge this represents the most comprehensive resour...

The routing of employees who provide services such as home health or social care is a complex problem. When sending an employee between two addresses, there may exist more than one travel option, e.g. public transport or car. In this paper we examine the optimisation of travel plans, supporting the provision of health and social services, with resp...

Human social hierarchy has the unique characteristic of existing in two forms. Firstly, as an informal hierarchy where leaders and followers are implicitly defined by their personal characteristics, and secondly, as an institutional hierarchy where leaders and followers are explicitly appointed by group decision. Although both forms can reduce the...

This paper tested the ability of machine learning techniques, namely artificial neural networks and random forests, to predict the individual trees within a forest most at risk of damage in storms. Models based on these techniques were developed individually for both a small forest area containing a set of 29 permanent sample plots that were damage...

Electronic institutions are socially-inspired multi-agent systems, typically operating under a set of policies, which are required to determine system operation and to deal with violations and other non-compliant behaviour. They are often faced with a dynamic population of agents, social network, and environment and their policy should suit this co...

One of the main components of multi-objective, and therefore, many-objective evolutionary algorithms, is the selection mechanism. It is responsible for performing two main tasks simultaneously. First, it has to promote convergence by selecting solutions which are as close as possible to the Pareto optimal set. And second, it has to promote diversit...

Workforce Scheduling and Routing Problems (WSRP) are very common in many practical domains, and usually have a number of objectives of interest to the end-user. Illumination algorithms such as Map-Elites (ME) have recently gained traction in application to design problems, in providing multiple diverse solutions as well as illuminating the solution...

Hyper-heuristic frameworks, although intended to be cross-domain at the highest level, rely on a set of domain-specific low-level heuristics at lower levels. For some domains, there is a lack of available heuristics, while for novel problems, no heuristics might exist. We address this issue by introducing a novel method, applicable in multiple doma...

The mutant vector generation strategy is an essential component of Differential Evolution (de), introduced to promote diversity, resulting in exploration of novel areas of the search space. However, it is also responsible for promoting intensification, to improve those solutions located in promising regions. In this paper we introduce a novel simil...

Hierarchy is an efficient way for a group to organize, but often goes along with inequality that benefits leaders. To control despotic behaviour, followers can assess leaders decisions by aggregating their own and their neighbours experience, and in response challenge despotic leaders. But in hierarchical social networks, this interactional justice...

Workforce Scheduling and Routing Problems (WSRP) are very common in many practical domains, and usually, have a number of objectives. Illumination algorithms such as Map-Elites (ME) have recently gained traction in application to {\em design} problems, in providing multiple diverse solutions as well as illuminating the solution space in terms of us...

The presence of functional diversity within a group has been demonstrated to lead to greater robustness, higher performance and increased problem-solving ability in a broad range of studies that includes insect groups, human groups and swarm robotics. Evolving group diversity however has proved challenging within Evolutionary Robotics, requiring re...

In many industrial problem domains, when faced with a combinatorial optimisation problem, a “good enough, quick enough” solution to a problem is often required. Simple heuristics often suffice in this case. However, for many domains, a simple heuristic may not be available, and designing one can require considerable expertise. Noting that a wide va...

Migrating Birds Optimisation (mbo) is a nature-inspired approach which has been shown to be very effective when solving a variety of combinatorial optimisation problems. More recently, an adaptation of the algorithm has been proposed that enables it to deal with continuous search spaces. We extend this work in two ways. Firstly, a novel leader repl...

Hierarchy is an efficient way for a group to organize, but often goes along with inequality that benefits leaders. To control despotic behaviour, followers can assess leaders’ decisions by aggregating their own and their neighbours’ experience, and in response challenge despotic leaders. But in hierarchical social networks, this interactional justi...

The sudden transition from egalitarian groups to hierarchical societies that occurred with the origin of agriculture is one of the most striking features of the evolution of human societies. Hierarchy is reflected by the evolution of an asymmetrical distribution of the influence of individuals. Although the benefits to leaders themselves are easily...

Selection methods are a key component of all multi-objective and, consequently, many-objective optimisation evolutionary algorithms. They must perform two main tasks simultaneously. First of all, they must select individuals that are as close as possible to the Pareto optimal front (convergence). Second, but not less important, they must help the e...

A robotic swarm that is required to operate for long periods in a potentially unknown environment can use both evolution and individual learning methods in order to adapt. However , the role played by the environment in influencing the effectiveness of each type of learning is not well understood. In this paper, we address this question by analysin...

Although the use of ensemble methods in machine-learning is ubiquitous, ensembles of optimisation algorithms have received relatively little attention. In [2] we address fundamental questions regarding ensemble composition in optimisation using the domain of bin-packing as a example. We first show that ensembles constructed by random selection from...

Recent research in small populations of Thymio II robots illustrated the relative benefits of populations distinguishing heritable and learning features in robots for a simple obstacle avoidance task. Here we scientifically validate these results by repeating them using a simulation. An additional benefit of this work is to provide confidence in th...

Catastrophic damage to forests resulting from major storms has resulted in serious timber and financial losses within the sector across Europe in the recent past. Developing risk assessment methods is thus one of the keys to finding forest management strategies to reduce future damage. Previous approaches to predicting damage to individual trees ha...

Although the use of ensemble methods in machine-learning is ubiquitous due to their proven ability to outperform their constituent algorithms, ensembles of optimisation algorithms have received relatively little attention. Existing approaches lag behind machine-learning in both theory and practice, with no principled design guide-lines available. I...

It is well known that in open-ended evolution, the nature of the environment plays in key role in directing evolution. However, in Evolutionary Robotics, it is often unclear exactly how parameterisation of a given environment might influence the emergence of particular behaviours. We consider environments in which the total amount of energy is para...

We present novel algorithmic schemes for dealing with large scale continuous problems. They are based on the recently proposed population-based meta-heuristics Migrating Birds Optimisation (MBO) and Multi-leader Migrating Birds Optimisation (MMBO), that have shown to be effective for solving combinatorial problems. The main objective of the current...

Hyper-heuristic methods for solving vehicle routing problems (VRP) have proved promising on a range of data. The vast majority of approaches apply selective hyper-heuristic methods that iteratively choose appropriate heuristics from a fixed set of pre-defined low-level heuristics to either build or perturb a candidate solution. We propose a novel h...

One of the important factors for any effective metaheuristic optimization algorithms is its ability to maintain a diverse population throughout the generations. Specifically, diversity maintenance can lead to exploration of novel areas of the search space, and hence potentially locate better solutions. Recently, a novel diversity technique called t...

Distributed autonomous systems consisting of large numbers of components with no central control point need to be able to dynamically adapt their controlmechanisms to deal with an unpredictable and changing environment. Existing frameworks for engineering self-adaptive systems fail to account for the need to incorporate self-expression - that is, t...

Choosing the correct algorithm to solve a problem still remains an issue 40 years after the Algorithm Selection Problem was first posed. Here we propose a hyper-heuristic which can apply one of two meta-heuristics at the current stage of the search. A scoring function is used to select the most appropriate algorithm based on an estimate of the impr...

We describe a new hyper-heuristic method NELLI-GP for solving job-shop scheduling problems (JSSP) that evolves an ensemble of heuristics. The ensemble adopts a divide-and-conquer approach in which each heuristic solves a unique subset of the instance set considered. NELLI-GP extends an existing ensemble method called NELLI by introducing a novel he...