Summary
Marlena Duda is a PhD-trained data scientist and postdoctoral researcher based in Atlanta with 9 years of hands-on experience applying machine learning to large-scale neuroimaging and clinical datasets. She develops ICA-based multimodal fusion methods and time-varying functional dynamics models for applications such as schizophrenia, neurodevelopment, and brain age prediction on cohorts exceeding 10,000 subjects. Her work has produced 11 peer-reviewed papers, multiple invited talks, and cross-disciplinary collaborations spanning academic and industry settings, including an ML internship at Genentech where she built interpretable pan-cancer models. Comfortable bridging deep learning, LSTMs/transformers, and statistical rigor from a dual PhD/statistics training, she focuses on interpretable, clinically-relevant models rather than black-box performance. Outside the lab she teaches fitness classes, reflecting a commitment to translating complex science into accessible practice and maintaining balance between analytic intensity and personal wellbeing. Her rare combination of large-cohort neuroimaging expertise and practical model interpretability makes her especially suited for translational AI in healthcare.
9 years of coding experience
10 years of employment as a software developer
Doctor of Philosophy - PhD, Bioinformatics, Doctor of Philosophy - PhD, Bioinformatics at University of Michigan
Bachelor of Science (BS), Biology/Biological Sciences, General, Bachelor of Science (BS), Biology/Biological Sciences, General at Northeastern University