Papers by Prof. Fernando Buarque
2019 IEEE Latin American Conference on Computational Intelligence (LA-CCI), 2019
This work is intended to assess the learning capability of an agent implemented with a Dueling Do... more This work is intended to assess the learning capability of an agent implemented with a Dueling Double Deep Q-Network in the problem of parameter control for Evolutionary and Swarm-based algorithms. The objective is to build a general parameter control method for these algorithms, that can be used for any Population Based Algorithm (PBA) to solve any numerical optimization problem, implemented for any computing platform, and is able to choose a good sequence of parameter values for the PBA, given a time budget constraint. For the experiments, an implementation of the Particle Swarm Optimization for CUDA devices was chosen as the PBA and a set of well-known highly complex numerical minimization problems were used for the benchmark. The experiments showed that the agent is clearly able to evolve from a completely random decision policy to a fitness-minimization-oriented policy for most of the functions.
Bookmarks Related papers MentionsView impact
2017 IEEE Symposium Series on Computational Intelligence (SSCI), 2017
Electricity consumption has increased all around the world in the last decades. This has caused a... more Electricity consumption has increased all around the world in the last decades. This has caused a rise in the use of fossil fuels and in the harming of the environment. In the past years the use of renewable energies and reduction of consumption has growth in order to deal with that problem. The change in the production paradigm led to an increasing search of ways to shorten consumption and adapt to the production. One of the solutions for this problem is to use Demand Response systems. Lighting systems have a major role in electricity consumption, so they are very suitable to be applied in a Demand Response system, optimizing their use. This optimization can be made in different ways being one of them by using a heuristic algorithm. This paper focuses on the use of Fish School Search algorithm to optimize a lighting system, in order to understand its capability of dealing with a problem of this nature and compare it with other algorithms to evaluate its performance.
Bookmarks Related papers MentionsView impact
Medical Engineering & Physics, 2020
Bookmarks Related papers MentionsView impact
Gait & posture, Jan 24, 2017
The conventional methods to assess human gait are either expensive or complex to be applied regul... more The conventional methods to assess human gait are either expensive or complex to be applied regularly in clinical practice. To reduce the cost and simplify the evaluation, inertial sensors and adaptive algorithms have been utilized, respectively. This paper aims to summarize studies that applied adaptive also called artificial intelligence (AI) algorithms to gait analysis based on inertial sensor data, verifying if they can support the clinical evaluation. Articles were identified through searches of the main databases, which were encompassed from 1968 to October 2016. We have identified 22 studies that met the inclusion criteria. The included papers were analyzed due to their data acquisition and processing methods with specific questionnaires. Concerning the data acquisition, the mean score is 6.1±1.62, what implies that 13 of 22 papers failed to report relevant outcomes. The quality assessment of AI algorithms presents an above-average rating (8.2±1.84). Therefore, AI algorithms ...
Bookmarks Related papers MentionsView impact
Journal of Intelligent & Fuzzy Systems, 2016
Bookmarks Related papers MentionsView impact
2007 International Joint Conference on Neural Networks, 2007
This work introduces a new method for surface reconstruction based on Growing Self-organizing Map... more This work introduces a new method for surface reconstruction based on Growing Self-organizing Maps, which learn 3D coordinates of each vertex in a mesh as well as they learn the topology of the input data set. Each map grows incrementally producing meshes of different resolutions, according to the application needs. Another highlight of the presented algorithm refers to the reconstruction
Bookmarks Related papers MentionsView impact
2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence), 2008
Bookmarks Related papers MentionsView impact
IEEE Transactions on Neural Networks, 2010
In this paper, we propose a new method for surface reconstruction based on growing self-organizin... more In this paper, we propose a new method for surface reconstruction based on growing self-organizing maps (SOMs), called growing self-reconstruction maps (GSRMs). GSRM is an extension of growing neural gas (GNG) that includes the concept of triangular faces in the learning algorithm and additional conditions in order to include and remove connections, so that it can produce a triangular two-manifold mesh representation of a target object given an unstructured point cloud of its surface. The main modifications concern competitive Hebbian learning (CHL), the vertex insertion operation, and the edge removal mechanism. The method proposed is able to learn the geometry and topology of the surface represented in the point cloud and to generate meshes with different resolutions. Experimental results show that the proposed method can produce models that approximate the shape of an object, including its concave regions, boundaries, and holes, if any.
Bookmarks Related papers MentionsView impact
Computerized Medical Imaging and Graphics, 2009
Bookmarks Related papers MentionsView impact
InTech eBooks, Mar 16, 2012
Bookmarks Related papers MentionsView impact
Bookmarks Related papers MentionsView impact
Lecture Notes in Computer Science, 2018
The intermittency of wind remains the greatest challenge to its large scale adoption and sustaina... more The intermittency of wind remains the greatest challenge to its large scale adoption and sustainability of wind farms. Accurate wind power predictions therefore play a critical role for grid efficiency where wind energy is integrated. In this paper, we investigate two hybrid approaches based on the genetic algorithm (GA) and particle swarm optimisation (PSO). We use these techniques to optimise an Adaptive Neuro-Fuzzy Inference system (ANFIS) in order to perform one-hour ahead wind power prediction. The results show that the proposed techniques display statistically significant out-performance relative to the traditional backpropagation least-squares method. Furthermore, the hybrid techniques also display statistically significant out-performance when compared to the standard genetic algorithm.
Bookmarks Related papers MentionsView impact
PLOS ONE, Apr 20, 2017
Bookmarks Related papers MentionsView impact
Institution of Engineering and Technology eBooks, Oct 9, 2018
In this chapter, several nature-inspired optimization algorithms are used to update finite-elemen... more In this chapter, several nature-inspired optimization algorithms are used to update finite-element models (FEMs) of structural systems. Usually, the numerical models of real mechanical structures, which are obtained by the FEM approach, give different results compared to the experimental measurements. The mismatch between numerical and experimental results is caused by the variability of the model parameters as well as the mathematical simplifications made during the modeling process. The procedure of correcting the numerical models is known as model updating where several model parameters are adjusted to minimize the error between the measurements and the numerical model. In this chapter, the model-updating procedure is defined as an optimization problem where several swarm intelligence algorithms: particle swarm optimization (PSO), ant colony optimization (ACO) and fish school search (FSS) algorithms are used to update the FEMs of two structural systems: A five degree of freedom (DOF) mass-spring system and an unsymmetrical H-shaped structure with real measurements. The results obtained in this study are compared with the results obtained by the genetic algorithm (GA). As a result, the updating procedures based on FSS, ACO and PSO algorithms give better results than the GA approach. Furthermore, the updating problem, in this chapter, is reformulated as a multiobjective (MO) problem, and a multiobjective PSO (MOPSO) algorithm was used to update the five DOF mass-spring system. The MOPSO algorithm shows promising result in model updating.
Bookmarks Related papers MentionsView impact
Springer eBooks, 2017
Bookmarks Related papers MentionsView impact
Lecture Notes in Computer Science, 2016
Large-scale and distributed software development initiatives demand a systematic testing process ... more Large-scale and distributed software development initiatives demand a systematic testing process in order to prevent failures. Significant amount of resources are usually allocated on testing. Like any development and designing task, testing activities have to be prioritised in order to efficiently validate the produced code. By using source code complexity measurement, Computational Intelligence and Image Processing techniques, this research presents a new approach to prioritise testing efforts on large-scale and distributed software projects. The proposed technique was validated by automatically highlighting sensitive code within the Linux device drivers source code base. Our algorithm was able to classify 3, 077 from 35, 091 procedures as critical code to be tested. We argue that the approach is general enough to prioritise test tasks of most critical large-scale and distributed developed software such as: Operating Systems, Enterprise Resource Planning and Content Management systems.
Bookmarks Related papers MentionsView impact
Bookmarks Related papers MentionsView impact
Swarm Intelligence (SI)-based metaheuristics are frequently used to solve complex optimization pr... more Swarm Intelligence (SI)-based metaheuristics are frequently used to solve complex optimization problems, which are too hard to be solved by classic exact algorithms. Inspired by nature, SI particles move through a search space in pursuit of good solutions. Even using SI, solving some large problems still takes a lot of time, e.g., due to the high number of dimensions and large search spaces. In order to overcome this, parallel implementations of SI algorithms have been investigated. They are typically based on low-level approaches for parallelism, such as MPI, OpenMP, and CUDA, which are tedious and error-prone to use. To overcome these issues, frameworks for high-level parallel programming such as the Muenster Skeleton Library (Muesli) can be used. We show how two SI algorithms, namely Particle Swarm Optimization (PSO) and Fish School Search (FSS), can be implemented in Muesli easily. Experimental results demonstrate the obtained performance and good scalability.
Bookmarks Related papers MentionsView impact
2020 IEEE Congress on Evolutionary Computation (CEC), 2020
Bookmarks Related papers MentionsView impact
2016 IEEE Symposium Series on Computational Intelligence (SSCI), 2016
Bookmarks Related papers MentionsView impact
Uploads
Papers by Prof. Fernando Buarque