Departamento de Lenguajes y Ciencias de la Computación, Universidad de Málaga. Spain
Abstract:In this work, we solve the problem of finding the best locations to place stations for depositing/collecting shared bicycles. To do this, we model the problem as the p-median problem, that is a major existing localization problem in optimization. The p-median problem seeks to place a set of facilities (bicycle stations) in a way that minimizes the distance between a set of clients (citizens) and their closest facility (bike station). We have used a genetic algorithm, iterated local search, particle swarm optimization, simulated annealing, and variable neighbourhood search, to find the best locations for the bicycle stations and study their comparative advantages. We use irace to parameterize each algorithm automatically, to contribute with a methodology to fine-tune algorithms automatically. We have also studied different real data (distance and weights) from diverse open data sources from a real city, Malaga (Spain), hopefully leading to a final smart city application. We have compared our results with the implemented solution in Malaga. Finally, we have analyzed how we can use our proposal to improve the existing system in the city by adding more stations.
Abstract:Federated learning is a training paradigm according to which a server-based model is cooperatively trained using local models running on edge devices and ensuring data privacy. These devices exchange information that induces a substantial communication load, which jeopardises the functioning efficiency. The difficulty of reducing this overhead stands in achieving this without decreasing the model's efficiency (contradictory relation). To do so, many works investigated the compression of the pre/mid/post-trained models and the communication rounds, separately, although they jointly contribute to the communication overload. Our work aims at optimising communication overhead in federated learning by (I) modelling it as a multi-objective problem and (II) applying a multi-objective optimization algorithm (NSGA-II) to solve it. To the best of the author's knowledge, this is the first work that \texttt{(I)} explores the add-in that evolutionary computation could bring for solving such a problem, and \texttt{(II)} considers both the neuron and devices features together. We perform the experimentation by simulating a server/client architecture with 4 slaves. We investigate both convolutional and fully-connected neural networks with 12 and 3 layers, 887,530 and 33,400 weights, respectively. We conducted the validation on the \texttt{MNIST} dataset containing 70,000 images. The experiments have shown that our proposal could reduce communication by 99% and maintain an accuracy equal to the one obtained by the FedAvg Algorithm that uses 100% of communications.
Abstract:This article presents the problem of locating electric vehicle (EV) charging stations in a city by defining the Electric Vehicle Charging Stations Locations (EV-CSL) problem. The idea is to minimize the distance the citizens have to travel to charge their vehicles. EV-CSL takes into account the maximum number of charging stations to install and the electric power requirements. Two metaheuristics are applied to address the relying optimization problem: a genetic algorithm (GA) and a variable neighborhood search (VNS). The experimental analysis over a realistic scenario of Malaga city, Spain, shows that the metaheuristics are able to find competitive solutions which dramatically improve the actual installation of the stations in Malaga. GA provided statistically the best results.
Abstract:This article introduces Random Error Sampling-based Neuroevolution (RESN), a novel automatic method to optimize recurrent neural network architectures. RESN combines an evolutionary algorithm with a training-free evaluation approach. The results show that RESN achieves state-of-the-art error performance while reducing by half the computational time.
Abstract:This work proposes a novel approach to evaluate and analyze the behavior of multi-population parallel genetic algorithms (PGAs) when running on a cluster of multi-core processors. In particular, we deeply study their numerical and computational behavior by proposing a mathematical model representing the observed performance curves. In them, we discuss the emerging mathematical descriptions of PGA performance instead of, e.g., individual isolated results subject to visual inspection, for a better understanding of the effects of the number of cores used (scalability), their migration policy (the migration gap, in this paper), and the features of the solved problem (type of encoding and problem size). The conclusions based on the real figures and the numerical models fitting them represent a fresh way of understanding their speed-up, running time, and numerical effort, allowing a comparison based on a few meaningful numeric parameters. This represents a set of conclusions beyond the usual textual lessons found in past works on PGAs. It can be used as an estimation tool for the future performance of the algorithms and a way of finding out their limitations.
Abstract:Recurrent neural networks (RNNs) are a powerful approach for time series prediction. However, their performance is strongly affected by their architecture and hyperparameter settings. The architecture optimization of RNNs is a time-consuming task, where the search space is typically a mixture of real, integer and categorical values. To allow for shrinking and expanding the size of the network, the representation of architectures often has a variable length. In this paper, we propose to tackle the architecture optimization problem with a variant of the Bayesian Optimization (BO) algorithm. To reduce the evaluation time of candidate architectures the Mean Absolute Error Random Sampling (MRS), a training-free method to estimate the network performance, is adopted as the objective function for BO. Also, we propose three fixed-length encoding schemes to cope with the variable-length architecture representation. The result is a new perspective on accurate and efficient design of RNNs, that we validate on three problems. Our findings show that 1) the BO algorithm can explore different network architectures using the proposed encoding schemes and successfully designs well-performing architectures, and 2) the optimization time is significantly reduced by using MRS, without compromising the performance as compared to the architectures obtained from the actual training procedure.
Abstract:In 2019, around 57\% of the population of the world has broadband access to the Internet. Moreover, there are 5.9 billion mobile broadband subscriptions, i.e., 1.3 subscriptions per user. So there is an enormous interconnected computational power held by users all around the world. Also, it is estimated that Internet users spend more than six and a half hours online every day. But in spite of being a great amount of time, those resources are idle most of the day. Therefore, taking advantage of them presents an interesting opportunity. In this study, we introduce JSDoop, a prototype implementation to profit from this opportunity. In particular, we propose a volunteer web browser-based high-performance computing library. JSdoop divides a problem into tasks and uses different queues to distribute the computation. Then, volunteers access the web page of the problem and start processing the tasks in their web browsers. We conducted a proof-of-concept using our proposal and TensorFlow.js to train a recurrent neural network that predicts text. We tested it in a computer cluster and with up to 32 volunteers. The experimental results show that training a neural network in distributed web browsers is feasible and accurate, has a high scalability, and it is an interesting area for research.
Abstract:Recurrent neural networks have demonstrated to be good at solving prediction problems. However, finding a network that suits a problem is quite hard because of their high sensitivity to the hyperparameter configuration. Automatic hyperparameter optimization methods help to find the most suitable configuration, but they are not extensively adopted because of their high computational cost. In this work, we study the use of the mean absolute error random sampling to compare multiple-hidden-layer architectures and propose an evolutionary strategy-based algorithm that uses its results to optimize the configuration of a recurrent network. We empirically validate our proposal and show that it is possible to predict and compare the expected performance of a hyperparameter configuration in a low-cost way, as well as use these predictions to optimize the configuration of a recurrent network.
Abstract:Cloud Computing is becoming the leading paradigm for executing scientific and engineering workflows. The large-scale nature of the experiments they model and their variable workloads make clouds the ideal execution environment due to prompt and elastic access to huge amounts of computing resources. Autoscalers are middleware-level software components that allow scaling up and down the computing platform by acquiring or terminating virtual machines (VM) at the time that workflow's tasks are being scheduled. In this work we propose a novel online multi-objective autoscaler for workflows denominated Cloud Multi-objective Intelligence (CMI), that aims at the minimization of makespan, monetary cost and the potential impact of errors derived from unreliable VMs. In addition, this problem is subject to monetary budget constraints. CMI is responsible for periodically solving the autoscaling problems encountered along the execution of a workflow. Simulation experiments on four well-known workflows exhibit that CMI significantly outperforms a state-of-the-art autoscaler of similar characteristics called Spot Instances Aware Autoscaling (SIAA). These results convey a solid base for deepening in the study of other meta-heuristic methods for autoscaling workflow applications using cheap but unreliable infrastructures.
Abstract:Deep learning hyper-parameter optimization is a tough task. Finding an appropriate network configuration is a key to success, however most of the times this labor is roughly done. In this work we introduce a novel library to tackle this problem, the Deep Learning Optimization Library: DLOPT. We briefly describe its architecture and present a set of use examples. This is an open source project developed under the GNU GPL v3 license and it is freely available at https://github.com/acamero/dlopt