Transferlernen in der Biomedizin

Abstract

Machine learning is commonly employed to extract meaningful information from large and complex data. In situations where only scantdata is available, algorithms can leverage abundant data from a separate (unrelated) context to address the learning problem. Here, we
present two recently developed biomedical applications that take
advantage of transfer learning to bridge the gap from model Systems to human: single-cell label transfer and drug response prediction inpatients.

Datei