Patrícia de Oliveira Lucas
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Patrícia de Oliveira Lucas has a total of 19 co-authors. Here are the top co-authors by number of shared publications:
- Frederico Gadelha Guimarães
Universidade Federal de Minas Gerais
8 shared publications - Petrônio Cândido Lima Silva
Federal Institute of Northern Minas Gerais, Januária, Brazil
6 shared publications - Eduardo Mazoni Andrade Marçal Mendes
Federal University of Minas Gerais
3 shared publications - Omid Orang
Federal University of Minas Gerais
3 shared publications - Ronaldo Medeiros dos Santos
Instituto Federal do Norte de Minas Gerais, Brazil
2 shared publications - Adriano Veloso
Universidade Federal de Minas Gerais
1 shared publication
Latest Publications
- 16 September 2025
- Published by Elsevier in Neurocomputing
- Vol. 656
Automated machine learning based on decomposition, causality and evolutionary multitask optimization for time series forecasting
- 16 September 2025
- Published by Elsevier in Neurocomputing
- Vol. 656
Automated machine learning based on decomposition, causality and evolutionary multitask optimization for time series forecasting
- Open Access
Causal Graph Fuzzy LLMs: A First Introduction and Applications in Time Series Forecasting
In recent years, the application of Large Language Models (LLMs) to time series forecasting (TSF) has garnered significant attention among researchers. This study presents a new frame of LLMs named CGF-LLM using GPT-2 combined with fuzzy time series (FTS) and causal graph to predict multivariate time series, marking the first such architecture in the literature. The key objective is to convert numerical time series into interpretable forms through the parallel application of fuzzification and causal analysis, enabling both semantic understanding and structural insight as input for the pretrained GPT-2 model. The resulting textual representation offers a more interpretable view of the complex dynamics underlying the original time series. The reported results confirm the effectiveness of our proposed LLM-based time series forecasting model, as demonstrated across four different multivariate time series datasets. This initiative paves promising future directions in the domain of TSF using LLMs based on FTS.
Open AccessCausal Graph Fuzzy LLMs: A First Introduction and Applications in Time Series Forecasting
In recent years, the application of Large Language Models (LLMs) to time series forecasting (TSF) has garnered significant attention among researchers. This study presents a new frame of LLMs named CGF-LLM using GPT-2 combined with fuzzy time series (FTS) and causal graph to predict multivariate time series, marking the first such architecture in the literature. The key objective is to convert numerical time series into interpretable forms through the parallel application of fuzzification and causal analysis, enabling both semantic understanding and structural insight as input for the pretrained GPT-2 model. The resulting textual representation offers a more interpretable view of the complex dynamics underlying the original time series. The reported results confirm the effectiveness of our proposed LLM-based time series forecasting model, as demonstrated across four different multivariate time series datasets. This initiative paves promising future directions in the domain of TSF using LLMs based on FTS.
- Open Access
- 14 February 2025
- Published by Instituto Mineiro de Gestão das Águas in Revista Mineira de Recursos Hídricos
- Vol. 6, e025001
Efeitos das mudanças climáticas na região norte de Minas Gerais
The Northern region of Minas Gerais has been experiencing climate phenomena that cause significant social disruptions. Global Climate Models (GCMs), through the Shared Socioeconomic Pathways (SSPs), are widely used by the Coupled Model Intercomparison Project (CMIP) to generate robust future climate projections for the Intergovernmental Panel on Climate Change (IPCC). This study aimed to evaluate the consequences on the Climatological Water Balance (CWB) in cultivation areas of 4 municipalities in the Northern region of Minas Gerais, between the years 2021 to 2060, using the global climate model First Institute of Oceanography Earth System Model (FIO-ESM) version 2.0, under the SSP2-4.5 and SSP5-8.5 scenarios. The results indicated a trend of increasing drought periods, along with greater variability in precipitation volumes across the region, as well as a rise in precipitation volumes and extreme weather events by the end of the analyzed period.
- 14 February 2025
- Published by Instituto Mineiro de Gestão das Águas in Revista Mineira de Recursos Hídricos
- Vol. 6, e025001
Open AccessEfeitos das mudanças climáticas na região norte de Minas Gerais
The Northern region of Minas Gerais has been experiencing climate phenomena that cause significant social disruptions. Global Climate Models (GCMs), through the Shared Socioeconomic Pathways (SSPs), are widely used by the Coupled Model Intercomparison Project (CMIP) to generate robust future climate projections for the Intergovernmental Panel on Climate Change (IPCC). This study aimed to evaluate the consequences on the Climatological Water Balance (CWB) in cultivation areas of 4 municipalities in the Northern region of Minas Gerais, between the years 2021 to 2060, using the global climate model First Institute of Oceanography Earth System Model (FIO-ESM) version 2.0, under the SSP2-4.5 and SSP5-8.5 scenarios. The results indicated a trend of increasing drought periods, along with greater variability in precipitation volumes across the region, as well as a rise in precipitation volumes and extreme weather events by the end of the analyzed period.
- 26 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Fuzzy Systems
- Vol. 32 (6) , 3703-3715
Multiple-Input–Multiple-Output Randomized Fuzzy Cognitive Map Method for High-Dimensional Time Series Forecasting
Fuzzy Cognitive Maps (FCMs) have demonstrated considerable success in time series forecasting and are adept at handling uncertainties and capturing the dynamics of complex systems. Nevertheless, challenges still remain in the handling of multivariate high-dimensional time series using a time-effective learning algorithm. This paper introduces MRHFCM, a new methodology for predicting high-dimensional time series in multiple-input multiple-output (MIMO) systems. MRHFCM represents a hybrid method that combines data embedding transformation, randomized high-order FCM (R-HFCM), and an echo state network (ESN). The core of MRHFCM involves a cascade of R-HFCMs termed the CR-HFCM model. Each CR-HFCM comprises three layers: the input layer, reservoir (internal layer), and output layer. Notably, only the output layer is trainable, employing the least squares minimization algorithm. The weights within each sub-reservoir are randomly chosen and remain unchanged throughout the training procedure. Three real-world high-dimensional datasets are utilized to assess the performance of the proposed MRHFCM method. The results obtained reveal that our approach outperforms some existing baseline and state-of-the-art machine learning and deep learning forecasting techniques in terms of both accuracy and parsimony.
- 26 March 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Fuzzy Systems
- Vol. 32 (6) , 3703-3715
Multiple-Input–Multiple-Output Randomized Fuzzy Cognitive Map Method for High-Dimensional Time Series Forecasting
Fuzzy Cognitive Maps (FCMs) have demonstrated considerable success in time series forecasting and are adept at handling uncertainties and capturing the dynamics of complex systems. Nevertheless, challenges still remain in the handling of multivariate high-dimensional time series using a time-effective learning algorithm. This paper introduces MRHFCM, a new methodology for predicting high-dimensional time series in multiple-input multiple-output (MIMO) systems. MRHFCM represents a hybrid method that combines data embedding transformation, randomized high-order FCM (R-HFCM), and an echo state network (ESN). The core of MRHFCM involves a cascade of R-HFCMs termed the CR-HFCM model. Each CR-HFCM comprises three layers: the input layer, reservoir (internal layer), and output layer. Notably, only the output layer is trainable, employing the least squares minimization algorithm. The weights within each sub-reservoir are randomly chosen and remain unchanged throughout the training procedure. Three real-world high-dimensional datasets are utilized to assess the performance of the proposed MRHFCM method. The results obtained reveal that our approach outperforms some existing baseline and state-of-the-art machine learning and deep learning forecasting techniques in terms of both accuracy and parsimony.
- 31 December 2023
- Published by Associacao Brasileira de Inteligencia Computacional - ABRICOM
Explorando causalidade na seleção de variáveis para previsão de séries temporais multivariadas
A previsão de séries temporais desempenha um papel fundamental em diversos setores, fornecendo insights valiosos para o planejamento e tomada de decisões em várias áreas de aplicação. A seleção de características em series temporais e, por sua vez, um componente crucial no pipeline de aprendizado de máquina. Ao selecionar cuidadosamente um conjunto apropriado de variáveis, os modelos de aprendizado de máquina podem aumentar a precisão da previsão, reduzir o tempo de processamento e melhorar a interpretabilidade. Neste estudo, comparamos o método de descoberta causal PCMCI (Peter and Clark Momentary Conditional Independence) com outros métodos de seleção de características não causais, como Correlação, Algoritmos Genéticos e regularização LASSO. Três conjuntos de dados do mundo real são usados nos experimentos de acordo com tres resultados principais: tamanho do gráfico gerado (numero de variáveis selecionadas), tempo de execução e precisão da previsão em diferentes horizontes de 1, 3 e 5 passos a frente. Os resultados mostram a capacidade do PCMCI em ` gerar modelos de previsão mais enxutos, proporcionando maior interpretabilidade e reduzindo a suscetibilidade ao sobre-ajuste.
- 31 December 2023
- Published by Associacao Brasileira de Inteligencia Computacional - ABRICOM
Explorando causalidade na seleção de variáveis para previsão de séries temporais multivariadas
A previsão de séries temporais desempenha um papel fundamental em diversos setores, fornecendo insights valiosos para o planejamento e tomada de decisões em várias áreas de aplicação. A seleção de características em series temporais e, por sua vez, um componente crucial no pipeline de aprendizado de máquina. Ao selecionar cuidadosamente um conjunto apropriado de variáveis, os modelos de aprendizado de máquina podem aumentar a precisão da previsão, reduzir o tempo de processamento e melhorar a interpretabilidade. Neste estudo, comparamos o método de descoberta causal PCMCI (Peter and Clark Momentary Conditional Independence) com outros métodos de seleção de características não causais, como Correlação, Algoritmos Genéticos e regularização LASSO. Três conjuntos de dados do mundo real são usados nos experimentos de acordo com tres resultados principais: tamanho do gráfico gerado (numero de variáveis selecionadas), tempo de execução e precisão da previsão em diferentes horizontes de 1, 3 e 5 passos a frente. Os resultados mostram a capacidade do PCMCI em ` gerar modelos de previsão mais enxutos, proporcionando maior interpretabilidade e reduzindo a suscetibilidade ao sobre-ajuste.
- 31 December 2022
- Published by Associacao Brasileira de Inteligencia Computacional - ABRICOM in Learning and Nonlinear Models
- Vol. 19 (2) , 29-50
A Tutorial on Fuzzy Time Series Forecasting Models: Recent Advances and Challenges
Time series forecasting is a powerful tool in planning and decision making, from traditional statistical models to soft computing and artificial intelligence approaches several methods have been developed to generate increasingly accurate forecasts. Fuzzy Time Series (FTS) methods have been introduced in the early 1990’s to handle data uncertainty and to undercome the statistical assumptions of linearity. Many studies have been reporting their good accuracy, simplicity, potential for interpretability and reduced computational complexity. This paper presents a tutorial for FTS methods. First, a review of the relevant literature is made, offering a foundation on the main concepts and FTS-based models for different time series and different types of forecasts. Then, the current challenges and possible solutions, are discussed alongside a timeline of the research developed in this area by the authors that aims at filling some of these gaps. Finally, a tutorial on the pyFTS library is presented. PyFTS is an open and free library coded in Python programming language that was developed by the MINDS Lab (Laboratory of Machine Intelligence and Data Science) and, also provides a set of transformation functions for pre-processing time series and a set of metrics and databases for benchmarking, in addition to implementing several FTS models in the literature.
- 31 December 2022
- Published by Associacao Brasileira de Inteligencia Computacional - ABRICOM in Learning and Nonlinear Models
- Vol. 19 (2) , 29-50
A Tutorial on Fuzzy Time Series Forecasting Models: Recent Advances and Challenges
Time series forecasting is a powerful tool in planning and decision making, from traditional statistical models to soft computing and artificial intelligence approaches several methods have been developed to generate increasingly accurate forecasts. Fuzzy Time Series (FTS) methods have been introduced in the early 1990’s to handle data uncertainty and to undercome the statistical assumptions of linearity. Many studies have been reporting their good accuracy, simplicity, potential for interpretability and reduced computational complexity. This paper presents a tutorial for FTS methods. First, a review of the relevant literature is made, offering a foundation on the main concepts and FTS-based models for different time series and different types of forecasts. Then, the current challenges and possible solutions, are discussed alongside a timeline of the research developed in this area by the authors that aims at filling some of these gaps. Finally, a tutorial on the pyFTS library is presented. PyFTS is an open and free library coded in Python programming language that was developed by the MINDS Lab (Laboratory of Machine Intelligence and Data Science) and, also provides a set of transformation functions for pre-processing time series and a set of metrics and databases for benchmarking, in addition to implementing several FTS models in the literature.
- Open Access
- 20 July 2021
- Published by EIA Energy in Agriculture in Energia na Agricultura
- Vol. 36 (1) , 58-71
VIABILIDADE ECONÔMICA DA ENERGIA FOTOVOLTAICA PARA IRRIGANTES NO NORTE DE MINAS E VALE DO JEQUITINHONHA
VIABILIDADE ECONÔMICA DA ENERGIA FOTOVOLTAICA PARA IRRIGANTES NO NORTE DE MINAS E VALE DO JEQUITINHONHA MARCELO ROSSI VICENTE1, TARLEI APARECIDO SANTOS2, THALLES LOIOLA DIAS1, PATRÍCIA DE OLIVEIRA E LUCAS1, RONALDO MEDEIROS DOS SANTOS1, CAIO VINICIUS LEITE3 1 Instituto Federal de Ensino, Ciência e Tecnologia do Norte de Minas Gerais – campus Salinas (IFNMG – campus Salinas), Fazenda Varginha, Rod. Salinas/Taiobeiras, Km 02, 39560-000, Salinas, Minas Gerais, Brasil. marcelo.vicente@ifnmg.edu.br, thallesloiola1@gmail.com, patricia.lucas@ifnmg.edu.br, ronaldo.medeiros@ifnmg.edu.br. 2 Departamento de Engenharia Cartográfica, Universidade Federal De Pernambuco (UFPE), Av. da Arquitetura, s/n, Cidade Universitária, 50740-550, Recife, Pernambuco, Brasil. tarleyengflorestal@gmail.com. 3 Instituto Federal de Ensino, Ciência e Tecnologia de Brasília – campus Planaltina (IFB - campus Planaltina), Rodovia DF – 128, km 21, Zona Rural, 73380-900, Planaltina, Distrito Federal, Brasil. caio.leite@ifb.edu.br RESUMO: Embora careçam de estudos de viabilidade, soluções baseadas em energia fotovoltaica têm se tornado muito atrativas para irrigantes, principalmente devido às constantes elevações nos custos de eletricidade. Com isso, objetivou-se realizar um estudo sobre a viabilidade econômica do uso da energia fotovoltaica para irrigação nas regiões do Vale do Jequitinhonha e Norte de Minas Gerais. Para a realização do trabalho utilizou-se a base de dados disponibilizada pelo Atlas Brasileira de Energia Solar e como índice de viabilidade econômica da energia fotovoltaica, o custo nivelado de energia (LCOE). Os valores de LCOE foram comparados com valores ponderados da tarifa de energia elétrica, em R$ kWh do Grupo Tarifário B2 Rural em diferentes cenários com jornadas de trabalho (noturna e diurna). Observou-se que quanto maior a radiação solar disponível e quanto maior a potência do sistema fotovoltaico, maior a tendência de viabilidade do investimento, em virtude da redução do LCOE. O uso de energia fotovoltaica é viável, com exceção das menores potências no Vale do Jequitinhonha e microrregiões limítrofes com o Norte de Minas, quando comparada ao uso exclusivo diurno da irrigação. O uso majoritário da tarifa com desconto noturno é uma alternativa à energia fotovoltaica quando se trata de viabilidade econômica. Palavras - chave: irradiação solar, custo nivelado de energia, tarifa de energia elétrica. ECONOMIC FEASIBILITY OF PHOTOVOLTAIC ENERGY FOR FARMERS IN NORTHERN MINAS GERAIS STATE AND JEQUITINHONHA VALLEY Although there is a lack of feasibility studies, solutions based on photovoltaic energy have become attractive for farmers due to the constant increases in electricity costs. Thus, this paper aimed to evaluate economic viability of photovoltaic energy for irrigation in the regions of Jequitinhonha Valley and the Northern Minas Gerais State. The database uses in this study came form the Brazilian Atlas of Solar Energy was used. The levelized cost of energy (LCOE) was used as an economic viability index for photovoltaic energy. The LCOE values were compared with the energy tariff weighted values, in R$ kWh, of the B2 Rural Tariff Group in different nights and day-working hours scenarios. It was observed that the greater the generation potential and the power of the photovoltaic system, the greater the trend of investment viability due to the LCOE reduction. The use of photovoltaic energy is feasible, except to lower powers in the Jequitinhonha Valley and Northern Minas Gerais State neighboring microregions, when compared to daytime irrigation. The mainly use of the nightly discounted energy tariff is an alternative to photovoltaic energy when it comes to economic viability. Keywords: Solar radiation, LCOE, Energy tariff
- 20 July 2021
- Published by EIA Energy in Agriculture in Energia na Agricultura
- Vol. 36 (1) , 58-71
Open AccessVIABILIDADE ECONÔMICA DA ENERGIA FOTOVOLTAICA PARA IRRIGANTES NO NORTE DE MINAS E VALE DO JEQUITINHONHA
VIABILIDADE ECONÔMICA DA ENERGIA FOTOVOLTAICA PARA IRRIGANTES NO NORTE DE MINAS E VALE DO JEQUITINHONHA MARCELO ROSSI VICENTE1, TARLEI APARECIDO SANTOS2, THALLES LOIOLA DIAS1, PATRÍCIA DE OLIVEIRA E LUCAS1, RONALDO MEDEIROS DOS SANTOS1, CAIO VINICIUS LEITE3 1 Instituto Federal de Ensino, Ciência e Tecnologia do Norte de Minas Gerais – campus Salinas (IFNMG – campus Salinas), Fazenda Varginha, Rod. Salinas/Taiobeiras, Km 02, 39560-000, Salinas, Minas Gerais, Brasil. marcelo.vicente@ifnmg.edu.br, thallesloiola1@gmail.com, patricia.lucas@ifnmg.edu.br, ronaldo.medeiros@ifnmg.edu.br. 2 Departamento de Engenharia Cartográfica, Universidade Federal De Pernambuco (UFPE), Av. da Arquitetura, s/n, Cidade Universitária, 50740-550, Recife, Pernambuco, Brasil. tarleyengflorestal@gmail.com. 3 Instituto Federal de Ensino, Ciência e Tecnologia de Brasília – campus Planaltina (IFB - campus Planaltina), Rodovia DF – 128, km 21, Zona Rural, 73380-900, Planaltina, Distrito Federal, Brasil. caio.leite@ifb.edu.br RESUMO: Embora careçam de estudos de viabilidade, soluções baseadas em energia fotovoltaica têm se tornado muito atrativas para irrigantes, principalmente devido às constantes elevações nos custos de eletricidade. Com isso, objetivou-se realizar um estudo sobre a viabilidade econômica do uso da energia fotovoltaica para irrigação nas regiões do Vale do Jequitinhonha e Norte de Minas Gerais. Para a realização do trabalho utilizou-se a base de dados disponibilizada pelo Atlas Brasileira de Energia Solar e como índice de viabilidade econômica da energia fotovoltaica, o custo nivelado de energia (LCOE). Os valores de LCOE foram comparados com valores ponderados da tarifa de energia elétrica, em R$ kWh do Grupo Tarifário B2 Rural em diferentes cenários com jornadas de trabalho (noturna e diurna). Observou-se que quanto maior a radiação solar disponível e quanto maior a potência do sistema fotovoltaico, maior a tendência de viabilidade do investimento, em virtude da redução do LCOE. O uso de energia fotovoltaica é viável, com exceção das menores potências no Vale do Jequitinhonha e microrregiões limítrofes com o Norte de Minas, quando comparada ao uso exclusivo diurno da irrigação. O uso majoritário da tarifa com desconto noturno é uma alternativa à energia fotovoltaica quando se trata de viabilidade econômica. Palavras - chave: irradiação solar, custo nivelado de energia, tarifa de energia elétrica. ECONOMIC FEASIBILITY OF PHOTOVOLTAIC ENERGY FOR FARMERS IN NORTHERN MINAS GERAIS STATE AND JEQUITINHONHA VALLEY Although there is a lack of feasibility studies, solutions based on photovoltaic energy have become attractive for farmers due to the constant increases in electricity costs. Thus, this paper aimed to evaluate economic viability of photovoltaic energy for irrigation in the regions of Jequitinhonha Valley and the Northern Minas Gerais State. The database uses in this study came form the Brazilian Atlas of Solar Energy was used. The levelized cost of energy (LCOE) was used as an economic viability index for photovoltaic energy. The LCOE values were compared with the energy tariff weighted values, in R$ kWh, of the B2 Rural Tariff Group in different nights and day-working hours scenarios. It was observed that the greater the generation potential and the power of the photovoltaic system, the greater the trend of investment viability due to the LCOE reduction. The use of photovoltaic energy is feasible, except to lower powers in the Jequitinhonha Valley and Northern Minas Gerais State neighboring microregions, when compared to daytime irrigation. The mainly use of the nightly discounted energy tariff is an alternative to photovoltaic energy when it comes to economic viability. Keywords: Solar radiation, LCOE, Energy tariff
- 1 October 2020
- Published by Elsevier in Computers and Electronics in Agriculture
- Vol. 177
Reference evapotranspiration time series forecasting with ensemble of convolutional neural networks
- 1 October 2020
- Published by Elsevier in Computers and Electronics in Agriculture
- Vol. 177
Reference evapotranspiration time series forecasting with ensemble of convolutional neural networks
- 12 March 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network and Service Management
- Vol. 17 (3) , 1309-1321
Distributed Evolutionary Hyperparameter Optimization for Fuzzy Time Series
Time series forecasting is an essential task in the management of Smart Cities and Smart Grids, becoming even more challenging when it needs to deal with big data time series. The development of highly accurate machine learning models is yet harder when considering the optimization of hyperparameters, which is an expensive computational task. To tackle these challenges this work proposes the Distributed Evolutionary Hyperparameter Optimization (DEHO) for the Weighted Multivariate Fuzzy Time Series method (WMVFTS), a simple and non-parametric forecasting method with high scalability and accuracy, comprising a sequential training and forecasting procedure and a MapReduce extension for distributed processing. The proposed methods were evaluated using a cluster with commodity hardware and two big time series, showing increasing speed up for training and test times as new CPU cores are added to cluster. Then the DEHO method was executed in the computational cluster, achieving fast convergence and feasible processing time and generating highly accurate WMVFTS models.
- 12 March 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Network and Service Management
- Vol. 17 (3) , 1309-1321
Distributed Evolutionary Hyperparameter Optimization for Fuzzy Time Series
Time series forecasting is an essential task in the management of Smart Cities and Smart Grids, becoming even more challenging when it needs to deal with big data time series. The development of highly accurate machine learning models is yet harder when considering the optimization of hyperparameters, which is an expensive computational task. To tackle these challenges this work proposes the Distributed Evolutionary Hyperparameter Optimization (DEHO) for the Weighted Multivariate Fuzzy Time Series method (WMVFTS), a simple and non-parametric forecasting method with high scalability and accuracy, comprising a sequential training and forecasting procedure and a MapReduce extension for distributed processing. The proposed methods were evaluated using a cluster with commodity hardware and two big time series, showing increasing speed up for training and test times as new CPU cores are added to cluster. Then the DEHO method was executed in the computational cluster, achieving fast convergence and feasible processing time and generating highly accurate WMVFTS models.
- 27 April 2019
- Published by Springer Nature
- p. 42-56
A Distributed Algorithm for Scalable Fuzzy Time Series
- 27 April 2019
- Published by Springer Nature
- p. 42-56
A Distributed Algorithm for Scalable Fuzzy Time Series