Examinando por Materia "Decision trees"
Mostrando 1 - 7 de 7
Resultados por página
Opciones de ordenación
Ítem Cerebral Cortex Atlas of Emotional States Through EEG Processing(SPRINGER, 2019-10-14) Gómez A.; Quintero O.L.; Lopez-Celani N.; Villa L.F.; Universidad EAFIT. Escuela de Ciencias; Modelado MatemáticoThis paper addresses the cerebral cortex maps construction from EEG signals getting an information simplification method for an emotional state phenomenon description. Bi-dimensional density distribution of main signal features are identified and a comparison to a previous approach is presented. Feature extraction scheme is performed via windowed EEG signals Stationary Wavelet Transform with the Daubechies Family (1–10); nine temporal and spectral descriptors are computed from the decomposed signal. Recursive feature selection method based on training a Random forest classifier using a one-vs-all scheme with the full features space, then a ranking procedure via gini importance, eliminating the bottom features and restarting the entire process over the new subset. Stopping criteria is the maximum accuracy. The main contribution is the analysis of the resulting subset features as a proxy for cerebral cortex maps looking for the cognitive processes understanding from surface signals. Identifying the common location of different emotional states in the central and frontal lobes, allowing to be strong parietal and temporal lobes differentiators for different emotions. © 2020, Springer Nature Switzerland AG.Ítem El incumplimiento crediticio mediante un análisis comparativo de diferentes técnicas de predicción en la geografía colombiana(Universidad EAFIT, 2021) Peñaloza Martínez, Eliana María; Tamara Ayús, Armando LeninÍtem Insolvencia empresarial en el sector salud colombiano(Universidad EAFIT, 2021) Echeverri Rentería, Nicolás; Tamara Ayus, Armando LeninÍtem Meteorological Risk Early Warning System for Air Operations(Institute of Electrical and Electronics Engineers Inc., 2019-01-01) Florez Zuluaga J.A.; David Ortega Pabon J.; Vargas Bonilla J.F.; Quintero Montova O.L.; Florez Zuluaga J.A.; David Ortega Pabon J.; Vargas Bonilla J.F.; Quintero Montova O.L.; Universidad EAFIT. Departamento de Ciencias; Modelado MatemáticoToday, airspace control has the challenge of merging information from independent and heterogeneous systems in order to minimize air safety risks and facilitate the decision-making process. One of the main risks for air operations is meteorology because convective formations like Torre cumulus or cumulonimbus could generate several dangerous phenomena such as icing, wind gusts, and thunderstorms, among others, that can affect the air operation safety. Based on previous works that allow the automatic identification of convective phenomena through the fusion of multispectral satellite images and other sources as winds and Meteorological Aerodrome Report (METAR), and establishing a common georeferenced coordinates system like WGS-84, for all sources, it can generate a system that could calculate early alerts about hazardous weather conditions in the aircrafts proximality for air traffic control system. For this, a meteorological analysis system can generate information about convective clouds calculating area, heights, temperatures, risk level and position of the meteorological formation. Parallelly the convective cloud is surrounded by optimal elliptical forms centered on the convective formation, generating a meteorological object. On the other hand, there is a system responsible for monitoring the information of the surveillance sensors. This system fused the air traffic sensors available like primary and secondary radar signals and ADS-B sensors in a unique WGS-84 coordinates system. Finally, in a georeferenced raster-Type graphing system or in a Geographic Information System (GIS), the meteorological and surveillance information is correlated projecting the track routes generates by air traffic system and traces generated by meteorological objects in order to establish times and high-risk areas, early. With this information, the Air Traffic Controller (ATC) system users, could minimize risk areas and reorganize the air traffic flow. This methodology then, would contribute to the decision-making process of ATC, facilitating the air flow reorganization and minimizing meteorological risks. For the development of this project a cooperative experimental methodology by subsystems was used. It was based on an operational knowledge and normal operating procedures of the Colombian Air Force, integrated with radar tracking technologies that implement decision trees. These alerts allow the air traffic controller to assess the risk and in accordance with the evaluation, if necessary, reorganize the air traffic flow for a specific area before the aircraft enter areas of bad weather mitigating the risks. © 2019 IEEE.Ítem Modelo de seguimiento de riesgo de crédito para el cliente independiente de una entidad financiera de Valle del Cauca(Universidad EAFIT, 2020) García Montealegre, Nathalia; Granja García, Vanessa Nathaly; Ospina Mejía, Jaime AlbertoThe purpose of this research is to propose a credit risk monitoring model for the independent profile client portfolio of a financial entity in Colombia´s department of Valle del Cauca, which provides information related to relevant variables associated to the analysis and management of the credit risk of this segment. The development of the monitoring model allows the grouping of the clients according to their characteristics and classifying them by their level of risk, predicting a possible default with more certainty and taking preventive actions against this fact, allowing in turn to comply with the risk policies defined in the Credit Risk Management System (Sistema de administración de riesgo de crédito, SARC), regulated in Colombia by the Superintendencia Financiera and based on the reference framework of the Basel Accords. The methodology used is that of data mining techniques such as clustering; the evaluation is carried out using decision trees; at the end the recommendations are presented.Ítem Spatial layout and surface reconstruction from omnidirectional images(Institute of Electrical and Electronics Engineers Inc., 2016-01-01) Posada, L.F.; Velasquez-Lopez, A.; Posada, L.F.; Velasquez-Lopez, A.; Universidad EAFIT. Departamento de Ciencias; Ciencias Biológicas y Bioprocesos (CIBIOP)This paper presents a spatial layout recovery approach from single omnidirectional images. Vertical structures in the scene are extracted via classification from heterogeneous features computed at the superpixel level. Vertical surfaces are further classified according to their main orientation by fusing oriented line features, floor-wall boundary features and histogram of oriented gradients (HOG) with a Random Forest classifier. Oriented line features are used to build an orientation map that considers the main vanishing points. The floor-wall boundary feature attempts to reconstruct the scene shape as if it were observed from a bird's-eye view. Finally, the HOG descriptors are aggregated per superpixel and summarize the gradient distribution at homogeneous appearance regions. Compared to existing methods in the literature which rely only on corners or lines, our method gains statistical support from multiple cues aggregated per superpixel which provide more robustness against noise, occlusion, and clutter. © 2016 IEEE.Ítem Spatial layout and surface reconstruction from omnidirectional images(Institute of Electrical and Electronics Engineers Inc., 2016-01-01) Posada, L.F.; Velasquez-Lopez, A.This paper presents a spatial layout recovery approach from single omnidirectional images. Vertical structures in the scene are extracted via classification from heterogeneous features computed at the superpixel level. Vertical surfaces are further classified according to their main orientation by fusing oriented line features, floor-wall boundary features and histogram of oriented gradients (HOG) with a Random Forest classifier. Oriented line features are used to build an orientation map that considers the main vanishing points. The floor-wall boundary feature attempts to reconstruct the scene shape as if it were observed from a bird's-eye view. Finally, the HOG descriptors are aggregated per superpixel and summarize the gradient distribution at homogeneous appearance regions. Compared to existing methods in the literature which rely only on corners or lines, our method gains statistical support from multiple cues aggregated per superpixel which provide more robustness against noise, occlusion, and clutter. © 2016 IEEE.