Vanda Lourenço

This paper evaluates robust Random Forests for genomic prediction under contamination, finding data transformation as the most effective strategy for improving performance in contaminated datasets.

This paper compares machine learning methods for genomic prediction using synthetic and empirical data, finding that no single method outperforms others consistently and that classical methods remain competitive.

The paper evaluates robust and classical methods for estimating heritability and predictive accuracy in plant breeding using simulations and empirical data, finding robust methods more reliable in contaminated data.