Buguemar
PhD student at Hasso Plattner Institut (HPI) - Artificial Intelligence and Intelligent Systems
HPIGermany
Pinned Repositories
buguemar
DataMining
Tareas_varias
graphing-a-decision
Public repository of our paper "Graphing a Decision: a Survey for Explainability on Graph-based Learning Models"
GRTC_GNNs
Public repository of our paper accepted to the Findings of EMNLP 2023: "Connecting the Dots: What Graph-Based Text Representations Work Best for Text Classification using Graph Neural Networks?"
ML
tareas ñanculef
NNTarea0
NNTarea3
PIIC19
Public repository of our works in Exoplanet analysis with Deep Learning
Poster
Auxiliar
Transformer_as_ensemble
Public repository of paper " Learning to combine classifiers outputs with the transformer for text classification"
Buguemar's Repositories
Buguemar/GRTC_GNNs
Public repository of our paper accepted to the Findings of EMNLP 2023: "Connecting the Dots: What Graph-Based Text Representations Work Best for Text Classification using Graph Neural Networks?"
Buguemar/Transformer_as_ensemble
Public repository of paper " Learning to combine classifiers outputs with the transformer for text classification"
Buguemar/buguemar
Buguemar/DataMining
Tareas_varias
Buguemar/graphing-a-decision
Public repository of our paper "Graphing a Decision: a Survey for Explainability on Graph-based Learning Models"
Buguemar/ML
tareas ñanculef
Buguemar/NNTarea0
Buguemar/NNTarea3
Buguemar/PIIC19
Public repository of our works in Exoplanet analysis with Deep Learning
Buguemar/Poster
Auxiliar
Buguemar/ProyectoIA-BEP
Proyecto final para Inteligencia Articial, Elizabeth Montero
Buguemar/ProyectoML
Poster, presentación magister
Buguemar/SIMAHcomp
Public repository of our 1st place work at the SIMAH competition held at ECML-PKDD 2019
Buguemar/tarea1_Redes
Buguemar/Tarea1ML
Desde cero juampi
Buguemar/Tarea2ML
Clasificadores en sklearn, fronteras, LDA/QDA/PCA, hiper parámetros.
Buguemar/Tarea3ML
Métodos No-Lineales
Buguemar/TEST
git commands
Buguemar/TransForE
Introducimos Transformer For Ensemble (TransForE), un método basado en Transformer para trabajar problemas de clasificación de texto de múltiples clases con un fuerte desequilibrio de etiquetas a fin de combinar el aprendizaje de múltiples modelos base a partir de las salidas de ellos, así como el texto mismo, en una especie de máquina de ensamblado parametrizada cuyo propósito es mejorar, o al menos mantener, la eficacia de los modelos base utilizados. TransForE utiliza los conocidos módulos de auto-atención de múltiples cabezales, propio de Transformer, con el propósito de aprender a combinar las múltiples componentes de entrada.