Sign Up to like & get
recommendations!
0
Published in 2025 at "Medical Physics"
DOI: 10.1002/mp.17834
Abstract: It is common in outcome analysis to work with a large set of candidate prognostic features. However, such high‐dimensional input and relatively small sample size leads to risk of overfitting, low generalizability, and correlation bias.
read more here.
Keywords:
clustering based;
approach address;
correlated features;
address correlated ... See more keywords