You can't ML your way out of bad data, no matter how many parameters you add.

Dr. Marc-André Schulz

Marc-André Schulz is a group leader in the Department of Machine Learning at the Hertie Institute for AI in Brain Health at the University of Tübingen. With a background in physics, Marc transitioned to machine learning and deep learning, specializing in the development and critical evaluation of machine learning methodologies for personalized psychiatry. Marc's research focuses on characterizing the constraints and limitations of these methods to assess if and under which conditions they offer potential for clinical translation.

You can't ML your way out of bad data, no matter how many parameters you add.

Siegel, N. T., Kainmueller, D., Deniz, F., Ritter, K., & Schulz, M. A.

Dr. Marc-André Schulz

Do Transformers and CNNs Learn Different Concepts of Brain Age?

DeepRepViz: Identifying Potential Confounders in Deep Learning Model Predictions

TLIMB-a transfer learning framework for image analysis of the brain

Performance reserves in brain-imaging-based phenotype prediction

Similar neural pathways link psychological stress and brain-age in health and multiple sclerosis

Data augmentation via partial nonlinear registration for brain-age prediction

A cognitive fingerprint in human random number generation

FIMAP: Feature Importance by Minimal Adversarial Perturbation

Promises and pitfalls of deep neural networks in neuroimaging-based psychiatric research

Different scaling of linear models and deep learning in UKBiobank brain images versus machine-learning datasets

Inferring disease subtypes from clusters in explanation space