Survival analysis for high-dimensional, heterogeneous medical data: exploring feature extraction as an alternative to feature selection

Conclusions If the number of samples is insufficient, feature extraction methods are unable to reliably identify the underlying manifold, which makes them of limited use in these situations. For large sample sizes – in our experiments, 2,500 samples or more – feature extraction methods perform as well as feature selection methods.
Source: Artificial Intelligence in Medicine - Category: Bioinformatics Source Type: research