17837 research outputs found
Sort by
Enhancing Quality-Diversity Optimization Through Domain-Specific Dissimilarity as Crowding Distance
Full text linkQuality-diversity algorithms aim to simultaneously optimize solution performance and maintain diversity within a population. In this paper, we explore the use of NSGA-II as a quality-diversity algorithm for the evolutionary design of 3D structures, modifying its crowding distance calculation to utilize dissimilarity measures. While NSGA-II is widely employed for multi-objective optimization, its use of fitness for calculating crowding distance may not be the most effective for tasks requiring solution diversity. We propose leveraging both genetic and phenotypic dissimilarity metrics to improve diversity management. To evaluate this approach, we compare the standard NSGA-II using fitness-based crowding distance and Diversity-Enhancing NSGA-II (DE-NSGA-II) using various combinations of dissimilarity-based metrics for crowding distance and diversity scores. Experiments are conducted using two distinct genetic representations on two optimization tasks: height of the center of gravity of passive structures and velocity of active structures. Results demonstrate the potential of dissimilarity-based crowding distance to enhance the diversity and overall quality of solutions in complex evolutionary design tasks
How Young People can Shape Environmental Policy in Urban Spaces
Full text linkYounger generations have become increasingly disillusioned with mainstream democratic politics in established democracies. Although young people are interested in politics and engaged in many issue-based forms of participation, it is hard for them to realise the fruits of their labour at the national level. Local democracy may provide a better opportunity for engaging effectively in the issues that affect young people’s everyday lives. This article examines how Public Value approaches work in practice for young people whose voices are usually excluded from the policy-making process. The research adopted a complex large-scale multi-stage qualitative design, that involved focus groups and interviews with young people and local civic leaders from across London. It used participatory research with young Londoners from traditionally marginalised groups. The research revealed that, although policy-makers face important structural challenges, such as the concentration of power and resources in Westminster, they have the potential to move beyond tokenistic engagement with young people. In particular, the results showed how civic and local authorities can build efficacy and trust through initiatives that provide opportunities for deliberation and the co-creation of public policy. In this way, the article makes a clear contribution to our understanding of the role of young people in environmentalism and their democratic value
DECML:Distributed Edge Consensus Machine Learning Framework
Full text linkThe increasing reliance on interdependent data-driven services in the Internet of Things (IoT), smart homes, and Industry 4.0 is hindered by siloed security and privacy measures. Existing solutions like federated learning and distributed machine learning, with their various approaches such as differential privacy and homomorphic encryption, while promising, face challenges in ensuring robust security and privacy. We introduce Distributed Edge Consensus Machine Learning (DECML), a novel framework that enables secure, privacy-preserving insights sharing among multiple stakeholders without exposing underlying data or models. DECML distributes queries and aggregates responses through independent nodes, achieving accuracy comparable to local deployments with minimal added latency. Our evaluation, using standard datasets and a 20-node network, demonstrates DECML's potential for collaborative decision-making without compromising privacy. This has significant implications for domains such as cybersecurity, healthcar
The Capacity of a Finite Field Matrix Channel
Full text linkThe Additive-Multiplicative Matrix Channel (AMMC) was introduced by Silva, Kschischang and Kötter in 2010 to model data transmission using random linear network coding. The input and output of the channel are matrices over a finite field . When the matrix is input, the channel outputs where is a uniformly chosen invertible matrix over and where is a uniformly chosen matrix over of rank .Silva et al. considered the case when . They determined the asymptotic capacity of the AMMC when , and are fixed and . They also determined the leading term of the capacity when is fixed, and , and grow linearly. We generalise these results, showing that the condition can be removed. (Our formula for the capacity falls into two cases, one of which generalises the case.) We also improve the error term in the case when is fixed
Auxiliary Generative Adversarial Networks with Iliustration2Vec and Q-Learning based Hyperparameter Optimisation for Anime Image Synthesis
Full text linkHarnessing the power of Generative Adversarial Networks (GANs) for the specialised task of anime face generation, this study introduces enhanced models of Auxiliary Classifier GAN (AC-GAN) and Wasserstein Auxiliary Classifier GAN (WAC-GAN) with modified network architectures and reinforcement learning-based hyperparameter optimisation. These models are uniquely adapted to handle the distinct nuances of anime-style imagery, a domain where conventional GANs often stumble due to complex stylistic variations and a heightened risk of mode collapse. Novel elements of our approach include, (1) modification of existing generator and discriminator architectures of both AC-GAN and WAC-GAN, (2) Q-learning based optimal hyperparameter selection, and (3) Illustration2Vec (I2V)-based automated attribute label extraction. Specifically, the Q-learning method is employed for hyperparameter search which effectively explores the search space of key network configurations by fulfilling the principles of Bellman optimality. Besides that, a deep learning-based 12V's method is utilised to generate attribute class labels and latent vectors to inform the generation process. Furthermore, we augment AC-GAN and WAC-GAN with additional layers to enhance their feature learning and generative capabilities. The insertion of these additional layers is calibrated based on the optimised network learning settings as well as the class labels derived from 12V, to fine-tune model scalability and diversity. Our experimental studies indicate that the conjunction of these techniques has led to a significant improvement in generating high-fidelity anime faces, adeptly handling the diverse and complex attributes inherent in anime-style imagery. The proposed strategies also showcase the potential of our customised AC-GAN and WAC-GAN models to master the nuanced art of anime face generation
Genetic Algorithm with Reinforcement Learning based Parameter Optimisation
Full text linkOptimisation of parameters in Genetic algorithms (GA) can improve the speed and accuracy of the solution produced, but well optimised parameters are dependant on the problem being solved, and the substantial additional cost of spending time pre-computing good parameters can offset the benefit. This research investigates the use of reinforcement learning algorithms to optimise the parameters of the GA during its runtime. Specifically, we propose a variant of the GA method which embeds the Q-learning algorithm to select an optimal mutation rate at each iteration. Evaluating with a set of benchmark functions, the proposed GA model with Q-learning shows promising performance with lower mean scores than those of the original GA for most test functions. In particular, the Q-learning algorithm shows a promising emergent behaviour, i.e. selecting a high mutation rate when the population variance is low to increase swarm and search diversity. Evaluated using diverse unimodal and multimodal numerical optimisation problems, the proposed model outperforms several baseline GAs with a statistical significance
Written evidence to the Northern Ireland Affairs Committee Inquiry on Policing and Security in Northern Ireland.
Full text linkThis submission provides responses to the Northern Ireland Affairs Committee’s ‘Policing and security in Northern Ireland’ inquiry evidence call. It focuses on two questions:a. What are the risks and opportunities associated with a process for the disbandment of paramilitary groups which the forthcoming independent scoping exercise should consider?b. What lessons should be learnt from previous attempts at paramilitary disbandment?<br/
Optimal Threshold Singular Spectrum Analysis for Efficient Electrocardiogram Interference Removal
Full text linkSingular Spectrum Analysis (SSA) has become well known for its ability to effectively separate mixtures of signals with overlapping spectral content but with different statistical natures. In this paper, we show how a new approach to grouping the singular values that efficiently denoise biomedical signals, specifically, mixtures of Electrocardiogram and Electromyogram signals. It is based on optimal Singular Value Hard Thresholding (SVHT) but for signals that are periodic or quasi-periodic in nature. An optimal thresholding technique can provide similar results with much smaller trajectory matrices and thus significantly reduced computational burden. The resultant Singular Value Decomposition process is significantly faster and shows similar performance to kurtosis based sliding SSA with a reduction in computational complexity of the order of 12,500 times. This technique is well suited to real-time implementation for de-noising biomedical signals on the fly
Data sources:
Access Repository Dashboard
Do you manage Royal Holloway - Pure? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!