Livia Moura Souza

This paper develops a machine learning approach for predicting rubber tree stem circumference using molecular markers, achieving higher accuracy than traditional models in a single-environment scenario.

This study characterizes the genetic diversity of 410 Brazilian grapevine accessions using microsatellite markers, identifies a core collection, and highlights applications for breeding programs.

This study compares genomic selection models for rubber trees, finding that multi-environment models with nonlinear and linear kernels improve prediction accuracy and genetic gain compared to single-environment models.