previous day
next day
all days

View: session overviewtalk overview

09:00-11:10 Session 8A: EvoMUSART 4 - Music & Sound
Location: Room A
Auralization of three-dimensional cellular automata
PRESENTER: Yuta Kariyado

ABSTRACT. An auralization tool for exploring three-dimensional cellular automata is presented. This proof-of-concept allows the creation of a sound field comprised of individual sound events associated with each cell in a three-dimensional grid. Each sound-event is spatialized depending on the orientation of the listener relative to the three-dimensional model. Conceived to be used as an immersive Virtual Reality (VR) scene, this software application also works as a regular application for environments where the VR infrastructure is missing. The proposed tool encourages the exploration of three-dimensional cellular automata, which is in general difficult, the discovery and storing of new patterns, and the use of these as generative music source.

The Simulated Emergence of Chord Function

ABSTRACT. In this paper, we propose an autonomous, unsupervised learning of chord classification, based on the Neural Hidden Markov Model (HMM), and extend it to the Semi-Makov Model (HSMM) to integrate such additional contexts as the pitch-class histogram, the beat positions, and the preceding chord sequences. We train our model on a minimally pre-processed dataset in a mixture of major/minor pieces, expecting the models to learn the chord clusters in accordance with the contexts without help of assignment of tonality. Thereafter, we evaluate their performance by perplexity, and show that the added contexts would considerably improve the efficiency. In addition, we show that the proposed model reflects the context of major and minor in its state transitions, even though trained in mixed tonality.

A Multi-Objective Evolutionary Approach to Identify Relevant Audio Features for Music Segmentation
PRESENTER: Igor Vatolkin

ABSTRACT. The goal of automatic music segmentation is to calculate boundaries between musical parts or sections that are perceived as semantic entities. Such sections are often characterized by specific musical properties such as instrumentation, dynamics, tempo, or rhythm. Recent data-driven approaches often phrase music segmentation as a binary classification problem, where musical cues for identifying boundaries are learned implicitly. Complementary to such methods, we present in this paper an approach for identifying relevant audio features that explain the presence of musical boundaries. In particular, we describe a multi-objective evolutionary feature selection strategy, which simultaneously optimizes two objectives. In a first setting, we reduce the number of features while maximizing an F-measure. In a second setting, we jointly maximize precision and recall values. Furthermore, we present extensive experiments based on six different feature sets covering different musical aspects. We show that feature selection allows for reducing the overall dimensionality while increasing the segmentation quality compared to full feature sets, with timbre-related features performing best.

SerumRNN: A White-box Audio VST Effect Programming System

ABSTRACT. Learning to program an audio production VST synthesizer is a time consuming process, usually obtained through inefficient trial and error and only mastered after years of experience. As an educational and creative tool for sound designers, we propose SerumRNN: a white-box, iterative system that provides step-by-step instructions for applying audio effects to change a user’s input audio towards a desired sound. We apply our system to Xfer Records Serum: currently one of the most popular and complex VST synthesizers used by the audio production community. Our results indicate that SerumRNN is consistently able to provide useful feedback for a variety of different audio effects and synthesizer presets. We demonstrate the benefits of using an iterative system and show that SerumRNN learns to prioritize effects and can discover more efficient effect order sequences than a variety of baselines.

Parameter tuning for wavelet-based sound event detection using Neural Networks
PRESENTER: Pallav Raval

ABSTRACT. Wavelet-based audio processing is used for sound event detection. The low-level audio features (timbral or temporal features) are found to be effective to differentiate between different sound event and that is why frequency processing algorithms have become popular in recent times. Wavelet based sound event detection is found effective to detect sudden onsets in audio signals because it offers unique advantages compared to traditional frequency-based sound event detection using CNN’s or RNN’s. In this work, wave-let transform is applied to the audio to extract audio features which can predict the occurrence of a sound event using a classical feedforward neural network. Additionally, this work attempts to identify the optimal wavelet parameters to enhance classification performance. 3 window sizes, 6 wavelet families, 4 wavelet levels, 3 decomposition levels and 2 classifier models are used for experimental analysis. The UrbanSound8k data is used and a classification accuracy up to 97% is obtained. Some major observations with regard to parameter-estimation are as follows: wavelet level and wavelet de-composition level should be low; it is desirable to have a large window; however, the window size is limited by the duration of the sound event. A window size greater than the duration of the sound event will decrease classification performance. All wavelet family can classify the sound events; however, using Symlet, Daubechies, Reverse biorthogonal and Biorthogonal families will save computational resources (lesser epochs) because they yield better accuracy compared to Fejér-Korovkin and Coiflets. This work conveys that wavelet-based sound event detection seems promising, and can be expanded to detect most of the common sounds and sudden events occurring at various environments.

09:00-11:10 Session 8B: EvoCOP 2
Location: Room B
An Artificial Immune System for Black Box Test Case Selection
PRESENTER: Lukas Rosenbauer

ABSTRACT. Testing is a crucial part of the development of a new product. For software validation a transformation from manual to automated tests can be observed which enables companies to implement large numbers of test cases. However, during testing situations may occur where it is not feasible to run all tests due to time constraints. Hence a set of critical test cases must be compiled which usually fulfills several criteria. Within this work we focus on criteria that are feasible for black box testing such as system tests. We adapt an existing artificial immune system for our use case and evaluate our method in a series of experiments using industrial datasets. We compare our approach with several other test selection methods where our algorithm shows superior performance.

On Hybrid Heuristics for Steiner Trees on the Plane with Obstacles
PRESENTER: Victor Parque

ABSTRACT. Minimal-length trees in the two-dimensional Euclidean domain are of special interest to enable the efficient coordination of multi-agent and interconnected systems. We propose an approach to compute obstacle-avoiding minimal trees by using the hybrid between hierarchical optimization of shortest routes through sequential quadratic programming over constrained two-dimensional convex domains, and the gradient-free stochastic optimization algorithms with a convex search space. Our computational experiments involving 3,000 minimal tree planning scenarios in maps with convex and non-convex obstacles show the feasibility and the efficiency of our approach. Also, our comparative study involving relevant classes of gradient-free and nature inspired heuristics has shed light on the robustness of the selective pressure and exploitation abilities of the Dividing Rectangles (DIRECT), the Rank-based Differential Evolution (RBDE) and the Differential Evolution with Successful Parent Selection (DESPS). Our approach offers the cornerstone mechanisms to further advance towards developing efficient network optimization algorithms with flexible and scalable representations.

Tabu-driven Quantum Neighborhood Samplers
PRESENTER: Charles Moussa

ABSTRACT. Combinatorial optimization is an important application tar- geted by quantum computing. However, near-term hardware constraints make quantum algorithms unlikely to be competitive when compared to high-performing classical heuristics on large practical problems. One option to achieve advantages with near-term devices is to use them in combination with classical heuristics. In particular, we propose using quantum methods to sample from classically intractable distributions – which is the most probable approach to attain a true provable quantum separation in the near-term – which are used to solve optimization prob- lems faster. We numerically study this enhancement by an adaptation of Tabu Search using the Quantum Approximate Optimization Algorithm (QAOA) as a neighborhood sampler. We show that QAOA provides a flexible tool for exploration-exploitation in such hybrid settings and can provide evidence that it can help in solving problems faster by saving many tabu iterations and achieving better solutions.

A Novel Ant Colony Optimization Strategy for the Quantum Circuit Compilation Problem
PRESENTER: Marco Baioletti

ABSTRACT. Quantum Computing represents the most promising technology towards speed boost in computation, opening the possibility of major breakthroughs in several disciplines including Artificial Intelligence. This paper investigates the performance of a novel Ant Colony Optimization (ACO) algorithm for the realization (compilation) of nearest-neighbor compliant quantum circuits of minimum duration. In fact, current technological limitations (e.g., decoherence effect) impose that the overall duration (makespan) of the quantum circuit realization be minimized, and therefore the production of minimum-makespan compiled circuits for present and future quantum machines is of paramount importance. In our ACO algorithm (QCC-ACO), we introduce a novel pheromone model, and we leverage a heuristic-based Priority Rule to control the iterative selection of the quantum gates to be inserted in the solution.

The proposed QCC-ACO algorithm has been tested on a set of quantum circuit benchmark instances of increasing sizes available from the recent literature. We demonstrate that the QCC-ACO obtains results that outperform the current best solutions in the literature against the same benchmark, succeeding in significantly improving the makespan values for a great number of instances and demonstrating the scalability of the approach.

09:00-11:10 Session 8C: EvoAPPS 5 - Best paper nominations
Location: Room C
EA-based ASV Trajectory Planner for Pollution Detection in Lentic Waters

ABSTRACT. This paper presents a new planner based on Evolutionary Algorithms (EAs) to optimize the trajectory of an Autonomous Surface Vehicle (ASV), equipped with a probe, that has to determine the location of a pollutant in lentic water bodies (e.g. reservoirs, dams). To achieve it, our planner 1) exploits the information provided by a simulator that determines the pollutant distribution based on the water currents and 2) is supported by an EA that optimizes the mission duration, the ASV trajectory length and the measurements taken by its probe in highly polluted areas. The current version of the planner also ensures that the trajectories are feasible from the ASV and water body perspective, and solves this constrained multi-objective problem as a mono-objective one that linearly combines the constraint and objective functions. The preliminary results over different scenarios show that the planner can already determine overall good solutions, but that needs to be modified (e.g. using a multi-objective intended EA) to improve them further.

Towards Feature-Based Performance Regression Using Trajectory Data
PRESENTER: Anja Jankovic

ABSTRACT. Black-box optimization is a very active area of research, with many new algorithms being developed every year. This variety is needed, on the one hand, since different algorithms are most suitable for different types of optimization problems. But the variety also poses a meta-problem: which algorithm to choose for a given problem at hand? Past research has shown that per-instance algorithm selection based on exploratory landscape analysis (ELA) can be an efficient mean to tackle this meta-problem. Existing approaches, however, require the approximation of problem features based on a significant number of samples, which are typically selected through uniform sampling or Latin Hypercube Designs. The evaluation of these points is costly, and the benefit of an ELA-based algorithm selection over a default algorithm must therefore be significant in order to pay off. One could hope to by-pass the evaluations for the feature approximations by using the samples that a default algorithm would anyway perform, i.e., by using the points of the default algorithm's trajectory. We analyze in this paper how well such an approach can work. Concretely, we test how accurately trajectory-based ELA approaches can predict the final solution quality of the CMA-ES after a fixed budget of function evaluations. We observe that the loss of trajectory-based predictions can be surprisingly small compared to the classical global sampling approach, if the remaining budget for which solution quality shall be predicted is not too large. Feature selection, in contrast, did not show any advantage in our experiments and rather led to worsened prediction accuracy. The inclusion of state variables of CMA-ES only has a moderate effect on the prediction accuracy.

A NEAT Visualisation of Neuroevolution Trajectories
PRESENTER: Stefano Sarti

ABSTRACT. NeuroEvolution of Augmenting Topologies (NEAT) is a system for evolving neural network topologies along with weights that has proven highly effective and adaptable for solving challenging reinforcement learning tasks. This paper analyses NEAT through the lens of Search Trajectory Networks (STNs), a recently proposed visual approach to study the dynamics of evolutionary algorithms. Our goal is to improve the understanding of neuroevolution systems. We present a visual and statistical analysis contrasting the behaviour of NEAT, with and without using the crossover operator, when solving the two benchmark problems outlined in the original NEAT article: XOR and double-pole balancing. Contrary to what is reported in the original NEAT article, our experiments without crossover perform significantly better in both domains.

Improving Search Efficiency and Diversity of Solutions in Multiobjective Binary Optimization by Using Metaheuristics plus Integer Linear Programming

ABSTRACT. Metaheuristics for solving multiobjective problems can provide an approximation of the Pareto front in a short time, but can also have difficulties finding feasible solutions in constrained problems. Integer linear programming solvers, on the other hand, are good at finding feasible solutions, but they can require some time to find and guarantee the efficient solutions of the problem. In this work we combine these two ideas to propose a hybrid algorithm mixing an exploration heuristic for multiobjective optimization with integer linear programming to solve multiobjective problems with binary variables and linear constraints. The algorithm has been designed to provide an approximation of the Pareto front that is well-spread throughout the objective space. In order to check the performance, we compare it with three popular metaheuristics using two benchmarks of multiobjective binary constrained problems. The results show that the proposed approach provides better performance than the baseline algorithms in terms of number of the solutions, hypervolume, generational distance, inverted generational distance, and the additive epsilon indicator.

11:30-13:15 Session 9A: EvoAPPS 6 - Games & Deep Bio-inspired Algorithms
Location: Room A
Combining Multi-objective Evolutionary Algorithms with deep generative models towards focused molecular design
PRESENTER: Tiago Sousa

ABSTRACT. Recent advances in applying deep generative learning to molecular design have led to a large number of novel approaches to the targeted generation of molecules towards specific features and applications. In this work, we expand on the latent space navigation approach, where molecules are optimized by operating in their latent representation inside a deep auto-encoder, by introducing multi-objective evolutionary algorithms (MOEAs) and benchmarking the proposed framework on several objectives from recent literature. By using several case studies from literature, we show that our proposed method is capable of controlling abstract chemical properties, is competitive with other state-of-the-art methods and is also able to perform relevant tasks such as optimizing a predefined molecule while maintaining a similarity threshold. Also, MOEAs allow to generate molecules with a good level of diversity, which is a desired feature.

A Profile-Based ‘GrEvolutionary’ Hearthstone Agent
PRESENTER: Antonio Mora

ABSTRACT. In the last few years, the Hearthstone AI international Competition has been gaining fame among the scientific community. Several different entries have been presented using varied approaches. One of the best, EVA, was focused on a Greedy approach combined with an Evolutionary Algorithm. However, almost all the proposals (including EVA) were designed to work in a general way, i.e. for any of the possible heroes. This generalisation presents a flaw, since the exclusive cards per hero are not really exploited, nor their potential different behaviour profiles. This paper follows a similar philosophy to EVA, also hybridizing Greedy + EA, but having in mind three different (and extended among the community) archetypes or profiles: Aggro, Control and Midrange. Thus, three different behaviours have been optimized indeed aiming to create a more specialized agent able to use an AI engine depending on the hero to play with. To prove the value of the approach several experiments have been conducted, comparing the evolved agents with EVA in many different matches using three different heroes. The results show an improvement over EVA for the three profile-based agents, as well as an excellent performance when combined with a MonteCarlo Tree Search approach.

Playing with Dynamic Systems - Battling Swarms in Virtual Reality

ABSTRACT. In this paper, we present a serious game with the goal to provide an engaging and immersive experience to foster the players’ understanding of dynamic networked systems. Confronted with attacking swarm networks, the player has to analyse their underlying network topologies and to systematically dismantle the swarms using a set of different weapons. We detail the game design, including the artificial intelligence of the swarm, the play mechanics and and the level designs. Finally, we conducted an analysis of the play performances of a test group over the course of the game which revealed a positive learning outcome.

EvoCraft: A New Challenge for Open-Endedness
PRESENTER: Djordje Grbic

ABSTRACT. This paper introduces EvoCraft, a framework for Minecraft designed to study open-ended algorithms. We introduce an API that provides an open-source Python interface for communicating with Minecraft to place and track blocks. In contrast to previous work in Minecraft that focused on learning to play the game, the grand challenge we pose hereis to automatically search for increasingly complex artifacts in an open-ended fashion. Compared to other environments used to study open-endedness, Minecraft allows the construction of almost any kind of structure, including actuated machines with circuits and mechanical components. We present initial baseline results in evolving simple Minecraft creations through both interactive and automated evolution. While evolution succeeds when tasked to grow a structure towards a specific target,it is unable to find a solution when rewarded for creating a simple machine that moves. Thus, EvoCraft offers a challenging new environment for automated search methods (such as evolution) to find complex artifacts that we hope will spur the development of more open-ended algorithms. A Python implementation of the EvoCraft framework is available at: github.com/real-itu/Evocraft-py.

Continuous Ant-Based Neural Topology Search

ABSTRACT. This work introduces a novel, nature-inspired neural archi-tecture search (NAS) algorithm based on ant colony optimization, Continuous Ant-based Neural Topology Search (CANTS), which utilizes synthetic ants that move over a continuous search space based on the densityand distribution of pheromones, is strongly inspired by how ants movein the real world. The paths taken by the ant agents through the search space are utilized to construct artificial neural networks (ANNs). This continuous search space allows CANTS to automate the design of ANNs of any size, removing a key limitation inherent to many current NAS algorithms that must operate within structures with a size predetermined by the user. CANTS employs a distributed asynchronous strategy whichallows it to scale to large-scale high performance computing resources,works with a variety of recurrent memory cell structures, and makes useof a communal weight sharing strategy to reduce training time. The pro-posed procedure is evaluated on three real-world, time series predictionproblems in the field of power systems and compared to two state-of-the-art algorithms. Results show that CANTS is able to provide improvedor competitive results on all of these problems, while also being easier touse, requiring half the number of user-specified hyper-parameters

EDM-DRL: Toward Stable Reinforcement Learning through Ensembled Directed Mutation

ABSTRACT. Deep reinforcement learning (DRL) has experienced tremendous growth in the past few years. However, training stability of agents continues to be an open research question. Here, the authors present Ensembled Directed Mutation of Deep Reinforcement Learning (EDM-DRL) - a hybridization of evolutionary computing (EC), ensemble learning, and DRL methods as a means of mitigating training instability in DRL agents. We show that our method trains more consistently than synchronous Advantage Actor Critic (A2C). We also show that by employing our novel mutation and ensemble methods, performance of DRL agents can be improved during test time without sacrificing training stability. Further, though a similar number of time steps are used, we show that the EDM-DRL algorithm uses a mere 1% or less of the network parameter updates used in A2C. Finally, we conduct an ablation study to identify components within the EDM-DRL algorithm responsible for highest contribution.

11:30-13:15 Session 9B: EvoCOP 3
Location: Room B
Flowshop NEH-based heuristic recommendation

ABSTRACT. Flowshop problems (FSPs) have many variants and a broad set of heuristics proposed in the literature to solve them. Therefore, choosing the best heuristic and its parameters for a given FSP instance can be very challenging for practitioners. Per-instance Algorithm Configuration (PIAC) approaches aim at recommending the best algorithm configuration for a particular instance problem. This paper presents a PIAC methodology for building models to automatically configure the Nawaz, Encore, and Ham (NEH) algorithm which proved to be a good choice in most FSP variants. We use irace to build the performance dataset (problem features - algorithm configuration), while training tree-based Decision Tree and Random Forest models to recommend NEH configurations on unseen problems of the test set. Results show that recommended heuristics have good performance, especially those recommended by random forest models considering parameter dependencies.

An Improvement Heuristic Based on Variable Neighborhood Search for a Dynamic Orienteering Problem
PRESENTER: Hoang Thanh Le

ABSTRACT. The Dynamic Orienteering Problem (DOP) is studied where nodes change their value over time. An improvement heuristic that is based on Variable Neighborhood Search is proposed for the DOP. The new heuristic is experimentally compared with two heuristics that are based on state-of-the-art algorithms for the static Orienteering Problem. For the experiments several benchmark instances are used as well as instances that are generated from existing road networks. The results show that the new heuristic outperforms the other heuristics with respect to several evaluation criteria and different measures for run time. An additional experiment shows that the new heuristic can be easily adapted to become a standalone algorithm that does not need given initial solutions. The standalone version obtains better results than two state-of-the-art algorithms.

A Heuristic Algorithm for School Bus Routing with Bus Stop Selection

ABSTRACT. In this paper a heuristic algorithm is proposed for a school bus routing problem which is formulated as a capacitated and time-constrained open vehicle routing problem with a homogeneous fleet and single loads. The algorithm determines the selection of bus stops from a set of potential stops, the assignment of students to selected bus stops, and the routes along the selected bus stops. Its goals are to minimize the number of buses used, the total route journey time and the student walking distances. It also aims at balancing route journey times between buses. The performance of the algorithm is evaluated on a set of twenty real-world problem instances and compared against solutions achieved by a mixed integer programming model. Reported results indicate that the heuristic algorithm finds high-quality solutions in very short amounts of computational time.

Hybrid Heuristic and Metaheuristic for Solving Electric Vehicle Charging Scheduling Problem
PRESENTER: Imene Zaidi

ABSTRACT. The electric vehicle (EV) charging scheduling problem has become a research focus to mitigate the impact of large-scale deployment of EV in the near future. One of the main assumptions in literature is that there are enough charging points (CP) in the charging station to meet all charging demands. However, with the deployment of EVs, this assumption is no longer valid. In this paper, we address the electric vehicle charging problem in a charging station with a limited number of heterogeneous CPs and a limited overall power capacity. Before arriving at the station, the EV drivers submit charging demands. Then, the scheduler reserves a suitable CP for each EV and allocates the power efficiently so that the final state-of-charge at the departure time is as close as possible to the requested state-of-charge. We present two variants of the problem: a constant output power model and a variable power model. To solve these problems, heuristic and simulated annealing (SA) combined with linear programming are proposed. Simulation results indicate that the proposed approaches are effective in terms of maximizing the state-of-charge by the departure time for each EV.

11:30-13:15 Session 9C: EvoAPPS 7 - Image Analysis & Signal Processing
Location: Room C
RDE-OP:A Region-based Differential Evolution Algorithm Incorporation Opposition-Based Learning for Optimising the Learning Process of Multi-Layer Neural Networks

ABSTRACT. Learning in multi-layer neural networks (MLNNs) involves finding appropriate weights and biases and is a challenging and important task since the performance of MLNNs is directly dependent on the weights. Conventional algorithms such as back-propagation suffer from difficulties including a tendency to get stuck in local optima. Population-based metaheuristic algorithms are used to address this issue. In this paper, we propose a novel learning approach, RDE-OP, based on differential evolution (DE) boosted by a region-based scheme and an opposition based strategy. DE is an efficient population-based metaheuristic algorithm which has shown satisfactory performance in solving optimisation problems. Our approach integrates two effective concepts with DE. First, we find some regions based on a clustering algorithm in the search space and select the cluster centre to represent the cluster. Then, an updating scheme is proposed to include them in the current population. In the next step, our proposed algorithm benefits from a quasi-opposition-based learning strategy for improved exploration of the search space. Experimental results on different datasets and in comparison with both conventional and population-based approaches convincingly indicate excellent performance of RDE-OP.

Lateralized Approach for Robustness Against Attacks in Emotion Categorization from Images

ABSTRACT. Deep learning has achieved a high classification accuracy on image classification tasks, including emotion categorization. However, deep learning models are highly vulnerable to adversarial attacks. Even a small change, imperceptible to a human (e.g. one-pixel attack), can decrease the classification accuracy of deep models. One reason could be their homogeneous representation of knowledge that considers all pixels in an image to be equally important is easily fooled. Enabling multiple representations of the same object, e.g. at the constituent and holistic viewpoints provides robustness against attacking a single view. This heterogeneity is provided by lateralization in biological systems. Lateral asymmetry of biological intelligence suggests heterogeneous learning of objects. This heterogeneity allows information to be learned at different levels of abstraction, i.e. at the constituent and the holistic level, enabling multiple representations of the same object. This work aims to create a novel system that can consider heterogeneous features e.g. mouth, eyes, nose, and jaw in a face image for emotion categorization. The experimental results show that the lateralized system successfully considers constituent and holistic features to exhibit robustness to unimportant and irrelevant changes in an image, demonstrating performance accuracy better than (or similar) to the deep learning system (VGG19). Overall, the novel lateralized method shows a stronger resistance to changes (10.86% - 47.72% decrease) than the deep model (25.15% - 83.43% decrease). The advances arise by allowing heterogeneous features, which enable constituent and holistic representations of image components.

EDA-based optimization of blow--off valve positions for centrifugal compressor systems

ABSTRACT. Designing actuators is an important part of automation technology and indispensable for the operation of plants in process industry. This also applies to valves which are important actuators of compressor system. Compressor systems have a high degree of complexity due to the interconnection of many different components. Often simulation environments are used to test already designed actuators. In this study we show how a digital twin of a compressor system in combination with an Estimation of Distribution (EDA) algorithm can be used to facilitate the valve design. In addition, the installation position in the plant is also determined in order to achieve a desired operating behaviour during an emergency shutdown.

Estimation of Grain-level Residual Stresses in a Quenched Cylindrical Sample of Aluminum Alloy AA5083 using Genetic Programming

ABSTRACT. Residual stresses are originated during fabrication processes of metallic materials and their study is important to prevent catastrophic failure during service of components. There are two main types of residual stresses, depending on the length scale; macroscopic and microscopic. We present an approach using genetic programming to obtain the micro residual stresses in grains of a quenched cylindrical sample of aluminium alloy AA5083. This alloy has a micro-structure of formed grains with different orientation and stress state. To obtain the micro residual stresses of each grain we estimate the total residual stresses values for every crystallographic orientation using information from electron back-scattered and neutron diffraction experiments. This information includes orientation maps of the normal section to the cylinder axes and the particular orientation and dimensions of every grain. We assume that the micro residual stresses of each grain can be expressed as a function of these variables and use genetic programming to find this expression.

3D-2D Registration using X-ray Simulation and CMA-ES

ABSTRACT. Radiographs of the hand are useful in diagnosing and staging diseases such as rheumatoid arthritis (RA) and other musculoskeletal diseases. Radiographs are projections of 3D anatomy, with useful information such as pose and pathology becoming lost in the process. We propose a 3D hand pose recovery method for radiographs of hands using a novel hybrid image registration method. Our pose recovery pipeline consists of aligning a simulated X-ray (digitally reconstructed radiograph) of an articulated phantom mesh model to a real hand radiograph using Covariance Matrix Adaptation Evolution Strategy. Early results demonstrate that our approach works well. Further inquiry is required to evaluate the applicability of our registration approach of other articulated musculoskeletal anatomy.

13:15-14:15Lunch Break
14:15-16:00 Session 10A: EvoMUSART 5 - Music & Sound
Location: Room A
SyVMO: Synchronous Variable Markov Oracle for modeling and predicting multi-part musical structures dependencies
PRESENTER: Nádia Carvalho

ABSTRACT. We present SyVMO, an algorithmic extension of the Variable Markov Oracle (VMO) algorithm, to model and predict multi-part dependencies from symbolic music manifestations. Our model has been implemented as a software application named INCITe for computer-assisted algorithmic composition, which learns musical structures from style-agnostic musical sources with variable amounts of data, represented as multiple viewpoints. To evaluate the SyVMO model within INCITe, we adopted the Creative Support Index survey and semi-structured interviews. Four expert composers participated in the evaluation adopting both personal and exogenous music corpus of variable size. The results suggested that INCITe shows great potential to support creative music tasks, namely in assisting the composition process, in particular because the use of SyVMO allowed for creating polyphonic suggestions from style-agnostic musical sources that also maintain a coherent melodic structure.

Sculpture Inspired Musical Composition, One Possible Approach
PRESENTER: Francisco Braga

ABSTRACT. In this paper, we present a system that takes a 3D model of a sculpture as starting point to compose music. Our approach does not consider the semantics of the sculpture but rather looks at it abstractly. The results were promising: the majority of the participants gave a classification of 4 out of 5 to the preferred interpretations of the compositions and related them to the respective sculpture. This is a step to a possible model for inspiration.

Raga Recognition in Indian Classical Music using Deep Learning
PRESENTER: Devansh Shah

ABSTRACT. Raga is central to Indian Classical Music in both Carnatic Music as well as Hindustani Music. The benefits of identifying raga from audio are related but not limited to the fields of Music Information Retrieval, content-based filtering, teaching-learning and so on. A deep learning and signal processing based approach is presented in this work in order to recognise the raga from the audio source using raw spectrograms. The proposed preprocessing steps and models achieved 98.98% testing accuracy on a subset of 10 ragas in CompMusic dataset. A thorough study on the effect of various hyperparameters, sound source separation and silent part removal is carried out and is listed with results and findings of the same. A discussion on the predictions and behavior of deep learning models on audio apart from the dataset is also carried out. This new approach yields promising results and gives real time predictions from raw audio or recordings.

Convolutional Generative Adversarial Network, via Transfer Learning, for Traditional Scottish Music Generation

ABSTRACT. The concept of a Binary Multi-track Sequential Generative Adversarial Network (BinaryMuseGAN) used for the generation of music has been applied and tested for various types of music. However, the concept is yet to be tested on more specific genres of music such as traditional Scottish music, for which extensive collections are not readily available. Hence exploring the capabilities of a transfer learning approach on these types of music is an interesting challenge for the methodology. The curated set of MIDI Scottish melodies was preprocessed in order to obtain the same number of tracks used in the BinaryMuseGAN model; converted into pianoroll format and then used as training set to fine tune a pretrained model, generated from the Lakh MIDI dataset. The results obtained have been compared with the results obtained by training the same GAN model from scratch on the sole Scottish music dataset. Results are presented in terms of variation and average performances achieved at different epochs for five performance metrics, three adopted from the Lakh dataset (qualified note rate, polyphonicity, tonal distance) and two custom defined to highlight Scottish music characteristics (dotted rhythm and pentatonic note). From these results, the transfer learning method shows to be more effective, with lower number of epochs, to converge stably and closely to the original dataset reference metrics values.

14:15-16:00 Session 10B: EvoCOP 4 Best Paper Nominations
Location: Room B
Decomposition-based Multi-objective Landscape Features and Automated Algorithm Selection
PRESENTER: Raphaël Cosson

ABSTRACT. Landscape analysis is of fundamental interest for improving our understanding on the behavior of evolutionary search, and for developing general-purpose automated solvers based on techniques from statistics and machine learning. In this paper, we push a step towards the development of a landscape-aware approach by proposing a set of landscape features for multi-objective combinatorial optimization, by decomposing the original multi-objective problem into a set of single-objective sub-problems. Based on a comprehensive set of bi-objective rMNK-landscapes and three variants of the state-of-the-art MOEA/D algorithm, we study the association between the proposed features, the global properties of the considered landscapes, and algorithm performance. We also show that decomposition-based features can be integrated into an automated approach for predicting algorithm performance and selecting the most accurate one on blind instances. In particular, our study reveals that such a landscape-aware approach is substantially better than the single best solver computed over the three considered MOEA/D variants

Hybridization of Racing Methods with Evolutionary Operators for Simulation Optimization of Traffic Lights Programs

ABSTRACT. In many real-world optimization problems, like the traffic light scheduling problem tackled here, the evaluation of candidate solutions requires the simulation of a process under various possible scenarios.

Thus, good solutions should not only achieve good objective function values, but they must be robust (low variance) across all different scenarios. Previous work has revealed the effectiveness of IRACE for this task. However, the operators used by IRACE to generate new solutions were designed for configuring algorithmic parameters, that have various data types (categorical, numerical, etc.). Meanwhile, evolutionary algorithms have powerful operators for numerical optimization, which could help to sample new solutions from the best ones found in the search. Therefore, in this work, we propose a hybridization of the elitist iterated racing mechanism of IRACE with evolutionary operators from differential evolution and genetic algorithms. We consider a realistic case study derived from the traffic network of Malaga (Spain) with 275 traffic lights that should be scheduled optimally. After a meticulous study, we discovered that the hybrid algorithm comprising IRACE plus differential evolution offers statistically better results than conventional algorithms and also improves travel times and reduces pollution.

Stagnation Detection with Randomized Local Search

ABSTRACT. Recently a mechanism called stagnation detection was proposed that automatically adjusts the mutation rate of evolutionary algorithms when they encounter local optima. The so-called SD-(1+1)EA introduced by Rajabi and Witt (GECCO 2020) adds stagnation detection to the classical (1+1)EA with standard bit mutation that flips each bit independently with some mutation rate and raises this rate when the algorithm is likely to have encountered local optima. In this paper, we investigate stagnation detection in the context of the k-bit flip operator of randomized local search that flips k bits chosen uniformly at random and let stagnation detection adjust the parameter k. We obtain improved runtime results compared to the SD-(1+1)EA amounting to a speed-up of up to e=2.71... Moreover, we propose additional schemes that prevent infinite optimization times even if the algorithm misses a working choice of k due to unlucky events. Finally, we present an example where standard bit mutation still outperforms the local k-bit flip with stagnation detection.

Symmetry Breaking for Voting Mechanisms
PRESENTER: Andrew Sutton

ABSTRACT. Recently, Rowe and Aishwaryaprajna [FOGA 2019] introduced a simple majority vote technique that efficiently solves \textsc{Jump} with large gaps, \textsc{OneMax} with large noise, and any monotone function with a polynomial-size image. In this paper, we identify a pathological condition for this algorithm: the presence of spin-flip symmetry. Spin-flip symmetry is the invariance of a pseudo-Boolean function to complementation. Many important combinatorial optimization problems admit objective functions that exhibit this pathology, such as graph problems, Ising models, and variants of propositional satisfiability. We prove that no population size exists that allows the majority vote technique to solve spin-flip symmetric functions with reasonable probability. To remedy this, we introduce a symmetry-breaking technique that allows the majority vote algorithm to overcome this issue for many landscapes. We prove a sufficient condition for a spin-flip symmetric function to possess in order for the symmetry-breaking voting algorithm to succeed, and prove its efficiency on generalized TwoMax and families of constructed 3-NAE-SAT and 2-XOR-SAT formulas. We also prove that it fails on the one-dimensional Ising model, and suggest different techniques for overcoming this. Finally, we present empirical results that explore the tightness of the runtime bounds and the performance of the technique on randomized satisfiability variants.

14:15-16:00 Session 10C: EvoAPPS 8 - Evolutionary Machine Learning (ii)
Location: Room C
Demonstrating the Evolution of GANs through t-SNE
PRESENTER: Victor Costa

ABSTRACT. Generative Adversarial Networks (GANs) are powerful generative models that achieved strong results, mainly in the image domain. However, the training of GANs is not trivial, presenting some challenges tackled by different strategies. Evolutionary algorithms, such as COEGAN, were recently proposed as a solution to improve the GAN training, overcoming common problems that affect the model, such as vanishing gradient and mode collapse. In this work, we propose an evaluation method based on t-distributed Stochastic Neighbour Embedding (t-SNE) to assess the progress of GANs and visualize the distribution learned by generators in training. We propose the use of the feature space extracted from trained discriminators to evaluate samples produced by generators and from the input dataset. A metric based on the resulting t-SNE maps and the Jaccard index is proposed to represent the model quality. Experiments were conducted to assess the progress of GANs when trained using COEGAN. The results show both by visual inspection and metrics that the Evolutionary Algorithm gradually improves discriminators and generators through generations, avoiding problems such as mode collapse.

Evaluating Models with Dynamic Sampling Holdout

ABSTRACT. Automated Machine Learning (Auto-ML) is a growing field where several techniques are being developed to address the question of how to automate the process of defining machine learning pipelines, using diverse types of approaches and with relative success, but still, being far from solved.

Among these still unsolved questions, the computational cost is one of the major issues. In this context, evaluating a model takes a lot of time and resources, and yet that is still a step that has not received much attention in the Auto-ML literature.

In this sense, this work revisits the Auto-CVE (Automated Coevolutionary Voting Ensemble) and proposes a new method for model evaluation: the dynamic sampling holdout. When compared to the regular Auto-CVE with cross-validation and the popular TPOT algorithm, Auto-CVE with dynamic holdout shows competitive results in both predictive performance and computing time.

On the Effects of Absumption for XCS with Continuous-Valued Inputs
PRESENTER: Alexander Wagner

ABSTRACT. The rule-based XCS Classifier System (XCS) aims at forming classifiers which are as general as possible to achieve an optimal performance level. A too high generalization pressure may lead to over-general classifiers degrading the performance of XCS. To date, no method exists for XCS for real-valued input spaces (XCSR) to handle over-general classifiers ensuring an accurate population. The Absumption mechanism and the Specify operator, both developed for XCS with binary inputs, provide a promising basis for over-generality handling in XCSR. This paper introduces adapted versions of Absumption and Specify by proposing different identification and specialization strategies for the application in XCSR. To determine their potential, the adapted techniques will be evaluated in different classification problems, i.e., common benchmarks and real-world data from the agricultural domain, and in a multi-step problem.

Transfer Learning for Automated Test Case Prioritization using XCSF
PRESENTER: Lukas Rosenbauer

ABSTRACT. With the rise of test automation, companies start to rely on large amounts of test cases. However, there are situations where it is unfeasible to perform every test case as only a limited amount of time is available. Under such circumstances a set of crucial tests has to be compiled. Recent research has shown that reinforcement learning methods such as XCSF classifier systems are well-suited for this task. This work investigates whether reusing knowledge of XCSF-based agents is beneficial for prioritizing test cases and subsequently selecting test suites in terms of performance. We developed a simplistic population transformation and evaluate it in a series of experiments. Our evaluation shows that XCSF may indeed benefit from transfer learning for this use case.

16:15-18:00 Session 11A: EvoMUSART 6 - Music & Sound
Location: Room A
Chord Embeddings: Analyzing What They Capture and Their Role for Next Chord Prediction and Artist Attribute Prediction
PRESENTER: Allison Lahnala

ABSTRACT. Natural language processing methods have been applied in a variety of music studies, drawing the connection between music and language. In this paper, we expand those approaches by investigating chord embeddings, which we apply in two case studies to address two key questions: (1) what musical information do they capture; (2) and how might musical applications benefit from them? In our analysis, we show that they capture similarities between chords that adhere to important relationships described in music theory. In the first case study, we demonstrate that using chord embeddings in a next chord prediction task yields predictions that more closely match those by experienced musicians. In the second case study, we show the potential benefits of using the representations in tasks related to musical stylometrics.

A Fusion of Deep and Shallow Learning to Predict Genres Based on Instrument and Timbre Features
PRESENTER: Igor Vatolkin

ABSTRACT. Deep neural networks have recently received a lot of attention and have very successfully contributed to many music classification tasks. However, they have also drawbacks compared to the traditional methods: a very high number of parameters, a decreased performance for small training sets, lack of model interpretability, long training time, and hence a larger environmental impact with regard to computing resources. Therefore, it can still be a better choice to apply shallow classifiers for a particular application scenario with specific evaluation criteria, like the size of the training set or a required interpretability of models. In this work, we propose an approach based on both deep and shallow classifiers for music genre classification: The convolutional neural networks are trained once to predict instruments, and their outputs are used as features to predict music genres with a shallow classifier. The results show that the individual performance of such descriptors is comparable to other instrument-related features and they are even better for more than half of 19 genre categories.

Genre Recognition from Symbolic Music with CNNs
PRESENTER: Edmund Dervakos

ABSTRACT. In this work we study the use of convolutional neural networks for genre recognition in symbolically represented music. Specifically, we explore the effects of changing network depth, width and kernel sizes while keeping the number of trainable parameters and each block's receptive field constant. We propose an architecture for handling MIDI data which makes use of multiple resolutions of the input, called MuSeReNet - Multiple Sequence Resolution Network. Through our experiments we significantly outperform the state-of-the-art for MIDI genre recognition on the topMAGD and MASD datasets.

Creating a Digital Mirror of Creative Practice

ABSTRACT. This paper describes an ongoing project to create a "digital mirror" to my practice as a composer of contemporary classical music; that is, a system that takes descriptions (in code) of aspects of that practice, and reflects them back as computer-generated realisations. The paper describes the design process of this system, explains how it is implemented, and gives some examples of the material that it generates. The paper further discusses some broader issues about the technological approach to building creative systems, in particular the advantages and disadvantages of building bespoke algorithms for generating creative content vs. the use of optimisation or learning from examples.

From Music to Image, a Computational Creativity Approach
PRESENTER: Luís Aleixo

ABSTRACT. In this paper we propose a possible approach for a cross-domain association between the musical and visual domains. We present a system that generates abstract images having as inspiration music files as the basis for the creative process. The system extracts available features from a MIDI music file given as input, associating them to visual characteristics, thus generating three different outputs. First, the Random and Associated Images - that result from the application of our approach considering different shape’s distribution - and second, the Genetic Image, that is the result of the application of one Genetic Algorithm that considers music and color theory while searching for better results. The results of our evaluation conducted through online surveys demonstrate that our system is capable of generating abstract images from music, since a majority of users consider the images to be abstract, and that they have a relation with the music that served as the basis for the association process. Moreover, the majority of the participants ranked highest the Genetic Image.

An Application for Evolutionary Music Composition using Autoencoders

ABSTRACT. This paper presents a new interactive application that can generate music according to a user's preferences inspired by the process of biological evolution. The application composes sets of songs that the user can choose from as a basis for the algorithm to evolve new music. By selecting preferred songs over successive generations, the application allows the user to explore an evolutionary musical space. The system combines autoencoder neural networks and evolution with human feedback to produce music. The autoencoder component is used to capture the essence of musical structure from a known sample of songs in a lower-dimensional space. Evolution is then applied over this representation to create new songs based upon previously generated songs the user enjoys. In this research, we introduce the application design and explore and analyse the autoencoder model. The songs produced by the application are also analysed to confirm that the underlying model has the ability to create a diverse range of music. The application can be used by composers working with dynamically generated music, such as for video games and interactive media.

16:15-18:00 Session 11B: EuroGP 4 Best Paper Nominations
Location: Room B
Regenerating Soft Robots through Neural Cellular Automata
PRESENTER: Kazuya Horibe

ABSTRACT. One of the key differences between biological and artificial systems is regeneration through the dynamic plasticity of living tissues. Lack of regeneration is a significant limitation for robots where robust operation outside controlled environments such as factory lines remains a challenge. To aid in addressing this gap, in this paper we develop an approach for simulated soft robots to regrow parts of their morphology when being damaged. Although numerical simulations using soft robots have played an important role in their design, evolving robots with regenerative capabilities have so far received comparable little attention. Here we propose a model for soft robots that regenerate through a neural cellular automata. Importantly, this approach only relies on local cell information to regrow damaged components, opening interesting possibilities for generation of physical regenerable soft robots in the future. Our approach allows simulated soft robots that are global damaged to partially regenerate back to their original morphology through local cell interactions alone and regain some of their ability to locomote. These results take a step towards equipping artificial systems with regenerative capacities and could potentially allow for more robust operations in a variety of situations and environments.

Exploring the automatic design of U-Nets applied to segmentation problems

ABSTRACT. A U-Net is a convolutional neural network mainly used for image segmentation domains such as medical image analysis. It consists of an U-shaped architecture which is able to learn features from the dataset by using information from previous layers to reconstruct the images. As other deep neural networks, the U-Net architecture influences the efficiency and accuracy of the network. In this work, we propose the use of a grammar-based evolutionary algorithm for the automatic design of deep neural networks for image segmentation tasks. The approach used is called Dynamic Structured Grammatical Evolution (DSGE), which employs a grammar to define the building blocks that are used to compose the networks, as well as the rules that help build them. We perform a set of experiments on the BSDS500 and ISBI12 datasets, designing networks tuned to image segmentation and edge detection. Subsequently, by using image similarity metrics, the results of our best performing networks are compared with the original U-Net. The results show that the proposed approach is able to design a network that is less complex in the number of trainable parameters, while also achieving slightly better results than the U-Net with a more consistent training.

Towards incorporating Human Knowledge in Fuzzy Pattern Tree Evolution
PRESENTER: Aidan Murphy

ABSTRACT. This paper shows empirically that Fuzzy Pattern Trees (FPT) evolved using Grammatical Evolution (GE), a system we call FGE, meet the criteria to be considered a robust Explainable Artificial Intelligence (XAI) system. Experimental results show FGE achieves competitive results against state of the art black box methods on a set of real world benchmark problems. Various selection methods were investigated to see which was best for finding smaller, more interpretable models and a human expert was recruited to test the interpretability of the models found and to give a confidence score for each model. Models which were deemed interpretable but not trustworthy by the expert were seen to be outperformed in classification accuracy by those interpretable models which were not, validating that FGE can be a powerful XAI technique.

Evolutionary Neural Architecture Search Supporting Approximate Multipliers
PRESENTER: Michal Pinos

ABSTRACT. There is a growing interest in automated neural architecture search (NAS) methods. They are employed to routinely deliver high-quality neural network architectures for various challenging data sets and reduce the designer’s effort. The NAS methods utilizing multi-objective evolutionary algorithms are especially useful when the objective is not only to minimize the network error but also to minimize the number of parameters (weights) or power consumption of the inference phase. We propose a multi-objective NAS method based on Cartesian genetic programming for evolving convolutional neural networks (CNN). The method allows approximate operations to be used in CNNs to reduce power consumption of a target hardware implementation. During the NAS process, a suitable CNN architecture is evolved together with approximate multipliers to deliver the best trade-offs between the accuracy, network size and power consumption. The most suitable approximate multipliers are automatically selected from a library of approximate multipliers. Evolved CNNs are compared with common human-created CNNs of a similar complexity on the CIFAR-10 benchmark problem.

16:15-18:00 Session 11C: EvoAPPS 9 - Evolutionary Machine Learning (iii)
Location: Room C
Towards Explainable Exploratory Landscape Analysis: Extreme Feature Selection for Classifying BBOB Functions
PRESENTER: Quentin Renau

ABSTRACT. Facilitated by the recent advances of Machine Learning~(ML), the automated design of optimization heuristics is currently shaking up evolutionary computation~(EC). Where the design of hand-picked guidelines for choosing a most suitable heuristic has long dominated research activities in the field, automatically trained heuristics are now seen to outperform human-derived choices even for well-researched optimization tasks. ML-based EC is therefore not any more a futuristic vision, but has become an integral part of our community.

A key criticism that ML-based heuristics are often faced with is their potential lack of explainability, which may hinder future developments. This applies in particular to supervised learning techniques which extrapolate algorithms' performance based on exploratory landscape analysis~(ELA). In such applications, it is not uncommon to use dozens of problem features to build the models underlying the specific algorithm selection or configuration task. Our goal in this work is to analyze whether this meany features are indeed needed. Using the classification of the BBOB test functions as testbed, we show that a surprisingly small number of features -- often less than four -- can suffice to achieve a 98\% accuracy. Interestingly, the number of features required to meet this threshold is found to decrease with the problem dimension. We show that the classification accuracy transfers to settings in which several instances are involved in training and testing. In the leave-one-instance-out setting, however, classification accuracy drops significantly, and the transformation-invariance of the features becomes a decisive success factor.

Desirable Objective Ranges in Preference-based Evolutionary Multiobjective Optimization

ABSTRACT. In this paper, we propose a preference-based Evolutionary Multiobjective Optimization algorithm, at which the preferences are given in the form of desirable ranges for the objective functions, i.e. by means of aspiration and reservation levels. The aspiration levels are values to be achieved by the objectives, while the reservation levels are objective values not to be worsen. In the algorithm proposed, the first generations are performed using a set of weight vectors to initially con- verge to the region of the Pareto optimal front associated with the point formed with the reservation levels. At a certain moment, these weights are updated using the nondominated solutions generated so far, to re- direct the search towards the region which contains the Pareto optimal solutions with objective values among the desirable ranges. To this aim, the remaining number of generations are run using the updated weight vectors and the point formed with the aspiration levels. The computational experiment shows the potential of our proposal in 2, 3 and 5- objective problems, in comparison to other state-of-the-art algorithms.

Adaptive Covariance Pattern Search

ABSTRACT. Pattern search is a family of single solution deterministic optimisation algorithms for numerical optimisation. Pattern search algorithms generate a new candidate solution by means of an archive of potential moves, named pattern. This pattern is generated by a basis of vectors that span the domain where the function to optimise is defined.

The present article proposes an adaptive implementation of pattern search that performs, at run-time, a fitness landscape analysis of the problem to determine the pattern and adapt it to the geometry of the problem. The proposed algorithm, called Adaptive Covariance Pattern Search (ACPS) uses at the beginning the fundamental orthonormal basis (directions of the variables) to build the pattern. Subsequently, ACPS saves the successful visited solutions, calculates the covariance matrix associated with these samples, and then uses the eigenvectors of this covariance matrix to build the pattern. ACPS is a restarting algorithm that at each restart recalculates the pattern that progressively adapts to the problem to optimise.

Numerical results show that the proposed ACPS appears to be a promising approach on various problems and dimensions.

Salp Swarm Optimization Search Based Feature Selection for Enhanced Phishing Websites Detection
PRESENTER: Ruba Abu Khurma

ABSTRACT. Internet-connected devices are increasing rapidly. This facilitates transferring most of the real-world transactions to the cyber world. It follows that eCrime is growing continuously. Phishing is a cyber-attack carried out by intruders. They aim to deceive the users of the Internet to achieve their malicious goals. Therefore, experts have developed different approaches to protect financial transactions and personal login information of the users. Their primary concern is to detect the security breaches for online use of the Internet channels (eg. emails, SMS, webpages, and social platforms). In this paper, we propose a new phishing detection system based on the Salp Swarm Algorithm (SSA). The main objective is to maximize the classification performance and minimize the number of features of the phishing system. Different transfer function (TF) families: S-TFs, V-TFs, X-TFs, U-TFs, and Z-TFs are used to convert the continuous SSA into binary. Sixteen different binary versions of the SSA algorithm are produced based on different TFs. A comparison analysis is performed to pick up the best binarization method. The phishing system is validated by comparing it with three state-of-the-art algorithms. The results show that BSSA with X-TFs achieved the best results in terms of the used evaluation measures.