Batch normalization followed by merging is powerful for phenotype prediction integrating multiple heterogeneous studies.
Heterogeneity in different genomic studies compromises the performance of machine learning models in cross-study phenotype predictions. Overcoming heterogeneity when incorporating different studies in terms of phenotype prediction is a challenging and critical step for developing machine learning al...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
Public Library of Science (PLoS)
2023-10-01
|
Series: | PLoS Computational Biology |
Online Access: | https://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1010608&type=printable |