Listar Tesinas Ciencia de la Computación por autor "Recamonde-Mendoza, Mariana"
Mostrando ítems 21-24 de 24
-
Stability and performance of machine learning feature selection methods : a case study on COVID-19 diagnosis datasets
Lampert, Ricardo Vedovelli (2022) [Tesinas de grado]Feature selection (FS) is defined as a preprocessing step that aims to find the smallest possible subset of the features that is enough to achieve the data analysis goal, in this case, the classification. The stability of ... -
A study on graph neural networks for classification tasks and model interpretability on genomic datasets
Melo, Lucas Lima de (2024) [Tesinas de grado]Recently, a few works have started proposing the use of graph neural networks (GNNs) to embed knowledge of gene interactions in machine learning models and thus produce more robust classifiers for genomic classification ... -
Use of embedding concatenation and ensemble to improve node classification on graphs
Mello, Ana Paula Carolino de Oliveira (2021) [Tesinas de grado]Artificial intelligence (AI) is a powerful tool that can be used in several different fields to solve many problems, and its use has been increasing every year. However, traditional machine learning (ML) algorithms have a ... -
Viés racial ou pequeno tamanho amostral? Investigando o impacto de disparidade racial em dados genômicos na análise de sobrevida em câncer
Castro, Daniel Matos de (2023) [Tesinas de grado]Este trabalho tem como objetivo investigar o impacto do viés racial nos dados ômicos sobre o desempenho de modelos preditivos com algoritmos de aprendizado de máquina (AM). Visamos analisar como o desbalanceamento entre ...