Frost forecast is an important issue in climate research because of its economic impact on severa... more Frost forecast is an important issue in climate research because of its economic impact on several industries. In this study, we propose GRAST-Frost, a graph neural network (GNN) with spatio-temporal architecture, which is used to predict minimum temperatures and the incidence of frost. We developed an IoT platform capable of acquiring weather data from an experimental site, and in addition, data were collected from 10 weather stations in close proximity to the aforementioned site. The model considers spatial and temporal relations while processing multiple time series simultaneously. Performing predictions of 6, 12, 24, and 48 h in advance, this model outperforms classical time series forecasting methods, including linear and nonlinear machine learning methods, simple deep learning architectures, and nongraph deep learning models. In addition, we show that our model significantly improves on the current state of the art of frost forecasting methods.
2015 30th IEEE/ACM International Conference on Automated Software Engineering Workshop (ASEW), 2015
This research indicates a novel approach of evolutionary multi-objective optimization algorithms ... more This research indicates a novel approach of evolutionary multi-objective optimization algorithms meant for integrating collective intelligence methods into the optimization process. The new algorithms allow groups of decision makers to improve the successive stages of evolution via users' preferences and collaboration in a direct crowdsourcing fashion. They can, also, highlight the regions of Pareto frontier that are more relevant to the group of decision makers as to focus the search process mainly on those areas. As part of this work we test the algorithms performance when face with some synthetic problem as well as a real-world case scenario.
ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal
This paper presents a multi-agent framework using Net- Logo to simulate human and collective beha... more This paper presents a multi-agent framework using Net- Logo to simulate human and collective behaviors during emergency evacuations. Emergency situation appears when an unexpected event occurs. In indoor emergency situation, evacuation plans defined by facility manager explain procedure and safety ways to follow in an emergency situation. A critical and public scenario is an airportwhere there is an everyday transit of thousands of people. In this scenario the importance is related with incidents statistics regarding overcrowding and crushing in public buildings. Simulation has the objective of evaluating building layouts considering several possible configurations. Agents could be based on reactive behavior like avoid danger or follow other agent, or in deliberative behavior based on BDI model. This tool provides decision support in a real emergency scenario like an airport, analyzing alternative solutions to the evacuation process.
2009 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology, 2009
Abstract In this paper, we present a context-aware architecture to facilitate adaptable web servi... more Abstract In this paper, we present a context-aware architecture to facilitate adaptable web services. In our approach, users interact with the system by means of a speech-based engine that allows using spontaneous speech to access the different services. These ...
... Multi-Agent System Virginia Fuentes, Nayat Sánchez-Pi, Javier Carbó and José M. Molina Univer... more ... Multi-Agent System Virginia Fuentes, Nayat Sánchez-Pi, Javier Carbó and José M. Molina University Carlos III of Madrid, Computer Science Department, Applied Artificial Intelligence Group (GIAA), Avda. Universidad Carlos III 22, 28270 Colmenarejo, Spain ...
2018 International Joint Conference on Neural Networks (IJCNN)
This paper puts forward a new text to tensor representation that relies on information compressio... more This paper puts forward a new text to tensor representation that relies on information compression techniques to assign shorter codes to the most frequently used characters. This representation is language-independent with no need of pretraining and produces an encoding with no information loss. It provides an adequate description of the morphology of text, as it is able to represent prefixes, declensions, and inflections with similar vectors and are able to represent even unseen words on the training dataset. Similarly, as it is compact yet sparse, is ideal for speed up training times using tensor processing libraries. As part of this paper, we show that this technique is especially effective when coupled with convolutional neural networks (CNNs) for text classification at character-level. We apply two variants of CNN coupled with it. Experimental results show that it drastically reduces the number of parameters to be optimized, resulting in competitive classification accuracy values in only a fraction of the time spent by one-hot encoding representations, thus enabling training in commodity hardware.
Frost forecast is an important issue in climate research because of its economic impact on severa... more Frost forecast is an important issue in climate research because of its economic impact on several industries. In this study, we propose GRAST-Frost, a graph neural network (GNN) with spatio-temporal architecture, which is used to predict minimum temperatures and the incidence of frost. We developed an IoT platform capable of acquiring weather data from an experimental site, and in addition, data were collected from 10 weather stations in close proximity to the aforementioned site. The model considers spatial and temporal relations while processing multiple time series simultaneously. Performing predictions of 6, 12, 24, and 48 h in advance, this model outperforms classical time series forecasting methods, including linear and nonlinear machine learning methods, simple deep learning architectures, and nongraph deep learning models. In addition, we show that our model significantly improves on the current state of the art of frost forecasting methods.
2015 30th IEEE/ACM International Conference on Automated Software Engineering Workshop (ASEW), 2015
This research indicates a novel approach of evolutionary multi-objective optimization algorithms ... more This research indicates a novel approach of evolutionary multi-objective optimization algorithms meant for integrating collective intelligence methods into the optimization process. The new algorithms allow groups of decision makers to improve the successive stages of evolution via users' preferences and collaboration in a direct crowdsourcing fashion. They can, also, highlight the regions of Pareto frontier that are more relevant to the group of decision makers as to focus the search process mainly on those areas. As part of this work we test the algorithms performance when face with some synthetic problem as well as a real-world case scenario.
ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal
This paper presents a multi-agent framework using Net- Logo to simulate human and collective beha... more This paper presents a multi-agent framework using Net- Logo to simulate human and collective behaviors during emergency evacuations. Emergency situation appears when an unexpected event occurs. In indoor emergency situation, evacuation plans defined by facility manager explain procedure and safety ways to follow in an emergency situation. A critical and public scenario is an airportwhere there is an everyday transit of thousands of people. In this scenario the importance is related with incidents statistics regarding overcrowding and crushing in public buildings. Simulation has the objective of evaluating building layouts considering several possible configurations. Agents could be based on reactive behavior like avoid danger or follow other agent, or in deliberative behavior based on BDI model. This tool provides decision support in a real emergency scenario like an airport, analyzing alternative solutions to the evacuation process.
2009 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology, 2009
Abstract In this paper, we present a context-aware architecture to facilitate adaptable web servi... more Abstract In this paper, we present a context-aware architecture to facilitate adaptable web services. In our approach, users interact with the system by means of a speech-based engine that allows using spontaneous speech to access the different services. These ...
... Multi-Agent System Virginia Fuentes, Nayat Sánchez-Pi, Javier Carbó and José M. Molina Univer... more ... Multi-Agent System Virginia Fuentes, Nayat Sánchez-Pi, Javier Carbó and José M. Molina University Carlos III of Madrid, Computer Science Department, Applied Artificial Intelligence Group (GIAA), Avda. Universidad Carlos III 22, 28270 Colmenarejo, Spain ...
2018 International Joint Conference on Neural Networks (IJCNN)
This paper puts forward a new text to tensor representation that relies on information compressio... more This paper puts forward a new text to tensor representation that relies on information compression techniques to assign shorter codes to the most frequently used characters. This representation is language-independent with no need of pretraining and produces an encoding with no information loss. It provides an adequate description of the morphology of text, as it is able to represent prefixes, declensions, and inflections with similar vectors and are able to represent even unseen words on the training dataset. Similarly, as it is compact yet sparse, is ideal for speed up training times using tensor processing libraries. As part of this paper, we show that this technique is especially effective when coupled with convolutional neural networks (CNNs) for text classification at character-level. We apply two variants of CNN coupled with it. Experimental results show that it drastically reduces the number of parameters to be optimized, resulting in competitive classification accuracy values in only a fraction of the time spent by one-hot encoding representations, thus enabling training in commodity hardware.
Uploads
Papers by Nayat Sanchez-Pi