A5005445784- Water Systems and Optimization
- Topic Modeling
- Fault Detection and Control Systems
- Natural Language Processing Techniques
- Advanced Control Systems Optimization
- Data Stream Mining Techniques
- Advanced Multi-Objective Optimization Algorithms
- Anomaly Detection Techniques and Applications
- Digital Imaging for Blood Diseases
- Spatial and Panel Data Analysis
- Machine Learning and Data Classification
- SARS-CoV-2 detection and testing
- Biosensors and Analytical Detection
- Advanced biosensing and bioanalysis techniques
- Advanced SAR Imaging Techniques
- Water Quality Monitoring Technologies
- Smart Grid Energy Management
- Clinical Laboratory Practices and Quality Control
- Evolutionary Algorithms and Applications
- Control Systems and Identification
- Neural Networks and Applications
- Advanced Control Systems Design
- Machine Learning and ELM
- Time Series Analysis and Forecasting
- COVID-19 diagnosis using AI
Pontifícia Universidade Católica do Paraná
2017-2024
Laboratório Nacional de Computação Científica
2024
Fundação Oswaldo Cruz
2024
Solstício Energia (Brazil)
2023
Universidade Federal do Rio Grande do Norte
2022
IBM Research - Brazil
2020-2021
Hôpital intercantonal de la Broye
2021
Short-term forecasting plays an important role in the economic area. Several studies have been carried, where models with good forecast capacity, focusing on accuracy or stability, were built. Modeling only one of these characteristics without other can lead to a model lower generalization capacity. To deal such situation, this study proposes ensemble (EM) one, two and three months ahead 60 kg corn bag prices received by producers state Parana (Brazil). Additionally, feature extraction means...
Abstract The coronavirus pandemic accentuated the need for molecular diagnostic tests. A technique highly used to this end is Polymerase Chain Reaction (PCR)—a sensitive and specific commonly as gold standard diagnostics. However, it demands trained personnel high-maintenance equipment relatively time-consuming. An alternative Loop-Mediated Isothermal Amplification (LAMP) technique, which doesn’t sample purification or expensive equipment, similar PCR when compared in sensitivity...
Abstract Climate trends and weather indicators are used in several research fields due to their importance statistical modeling, frequently as covariates. Usually, climate available grid files with different spatial time resolutions. The availability of a series compatible administrative boundaries is scattered Brazil, not fully for years, produced diverse methodologies. In this paper, we propose the Brazilian municipalities using zonal statistics derived from ERA5-Land reanalysis...
Abstract The complete blood count (CBC) is one of the most requested tests by physicians. CBC tests, realized in conventional hematological analyzers, are restricted to centralized laboratories due frequent maintenance, large devices, and expensive costs required. On other hand, handheld devices commercially available show high prices not liable calibration or control procedures, which results poor quality compared standard hematology instruments. Hilab system a small-handed platform that...
This paper explores how intent classification can be improved by representing the class labels not as a discrete set of symbols but space where word graphs associated to each are mapped using typical graph embedding techniques. The approach, inspired previous algorithm used for an inverse dictionary task, allows take in account inter-class similarities provided repeated occurrence some words training examples different classes. is carried out mapping text embeddings Focusing solely on...
This paper reports a rapid and sensitive sensor for the detection quantification of COVID-19 N-protein (N-PROT) via an electrochemical mechanism. Single-frequency impedance spectroscopy was used as transduction method real-time measurement N-PROT in immunosensor system based on gold-conjugate-modified carbon screen-printed electrodes (Cov-Ag-SPE). The presents high selectivity attained through optimal stimulation signal composed 0.0 V DC potential 10 mV RMS−1 AC at 100 Hz over 300 s....
Machine learning algorithms have found to be useful for the solution of complex engineering problems. However, due problem's characteristics, such as class imbalance, classical methods may not formidable. The authors believe that application multi-objective optimization design can improve results machine on scenarios. Thus, this paper proposes a novel methodology creation ensembles classifiers. To do so, approach composed two steps is used. first step focus generating set diverse...
Renewable energy is changing the market environment in all countries. To keep competitiveness and attractiveness of sector, countries are optimizing their regulatory framework to absolve those changes. Distributed Energy Resources (DERs) one key aspects changes, empowering consumers taking meet necessities a higher level. In this context, Brazil discussing full opening its market, including residential consumers. The increasing participation DER (currently representing almost 5% power...
This document proposes the use of multi-objective machine learning in order to solve problem online anomaly detection for drinking water quality. Such consists an imbalanced data set where events, minority class, must be correctly detected based on a time series denoting quality and operative data. In develop two different robust systems, signal processing feature engineering are used prepare data, while evolutionary optimization is selection ensemble generation. The proposed systems tested...
Machine learning algorithms are valuable tools for solving a wide variety of complex engineering problems. Usually, those problems have multiple criteria to fulfill, but such machine learning-based solutions usually optimized using single criterion. In instances, multi-objective optimization-based approach could bring interesting by determining set Pareto-optimal with different trade-off. Therefore, multi-criteria decision-making process must be carried out. To the authors’ present...
Real-world classification problems generally deal with imbalanced data, where one class represents the majority of data set. The present work deals event detection on a drinking-water quality time series, presence is minority class. In order to solve such problems, supervised learning algorithms are recommended. Researchers have also used multi-objective optimization (MOO) in generate diverse models build ensembles classifiers. Although MOO has been for ensemble member generation, there lack...
Control of industrial plants is an important engineering field study, where the optimization such processes can lead to better product quality and higher profit. The present work deals with gasifier, proposed as a benchmark challenge by ALSTOM Power Technologies in 2002. Researchers from around globe different methods for optimizing system, using linear nonlinear system identification techniques, controller schemes, proportional-integral-derivative model predictive controllers, search...
This paper proposes the use of multi-objective ensemble learning to monitor drinking-water quality. Such problem consists a data set with an extreme imbalance ratio where events, minority class, must be correctly detected given time series denoting water quality and operative on minutely basis. First, is preprocessed for imputing missing data, adjusting concept drift adding new statistical features, such as moving average, standard deviation, maximum minimum. Next, two techniques are used,...
Dealing with real world engineering problems, often comes facing multiple and conflicting objectives requirements. Water distributions systems (WDS) are not exempt from this: while cost hydraulic performance usually objectives, several requirements related environmental issues in water sources might be conflict as well. Commonly, optimisation statements defined order to address the WDS design, management and/or control. Multi-objective can handle such by means of a simultaneous design...
Within unmanned aerial vehicles on-going research topics, automatic target recognition is acquiring relevance. This due to the easiness with which it possible acquire such devices. In order do this, signal processing and classification techniques could be adopted. Also, in improve ratio, optimization used. subject object of study for many researchers, but authors worry about lack information while using single-objective techniques. The proposed work comprises application multi-objective...
Within unmanned aerial vehicles on-going research topics, automatic target recognition is acquiring relevance. This due to the easiness with which it possible acquire such devices. To do so, signal processing and classification techniques could be used. Many can used, literature usually rely on use of time-frequency distributions. But, since there a high computational cost involved, authors worry about mathematical techniques, low computationally expensive technique applied create an system....
This paper describes how machine learning training data and symbolic knowledge from curators of conversational systems can be used together to improve the accuracy those enable better curatorial tools. is done in context a real-world practice who often embed taxonomically-structured meta-knowledge into their documentation. The provides evidence that quite common among curators, as part collaborative practices, embedded mined by algorithms. Further, this integrated, using neuro-symbolic...